UN Transcripts — https://transcripts.un.org/en/ecosoc/2026/12
(Continued) Special meeting on international cooperation in tax matters - Economic and Social Council, 12th plenary meeting, 2026 session — Economic and Social Council — 27 March 2026
Language: en
Automatically generated transcript — may contain errors. Not an official United Nations record.

---
ECOSOC · President [0:04]:

Good afternoon,.

Declaro Avierta. I hereby declare open the 12th session of the Economic and Social Council.

I now invite the Council to continue its consideration of sub item H of Agenda Item 18, entitled International Cooperation in Tax Matters for the second part of the Special Meeting on International Cooperation in Tax Matters. I would now like to welcome you to the continuation of the second panel discussion on permanent establishment and significant economic presence, choices, challenges and country experiences. I am pleased to welcome back our distinguished presenters and moderator for this discussion. We will now continue with the list of speakers from the floor. I would like to remind Legations that all of those that have signed up on the list have five minutes for statements on behalf of groups and three minutes for individual statements.

Once again, I kindly remind speakers to please deliver your statements at a normal speaking speed to facilitate interpretation. Please note that given the limited amount of time available, the microphone will be automatically cut off when the allotted time has elapsed. This will allow us to hear all of the speakers on the list. I now give the floor to the distinguished representative of Zimbabwe and they will be followed by the distinguished delegation of Algeria. Vice President of ECOSOC thank you, Mr.


Zimbabwe [2:03]:

President, for giving me the floor. We thank the panelists for their valuable insights. The discussion on permanent establishment and significant economic presence ultimately comes down to a fundamental issue. Where should taxing rights lie in an increasingly digital and globalized economy? For Zimbabwe, this is a practical and immediate concern.

A growing share of economic activity within our jurisdiction is generated by multinational enterprises, often through subsidiaries, but increasingly through remote and digital business models that do not require physical presence. This challenges the continued reliance on traditional permanent establishment rules, which were designed for a different economic reality. As a result, there is a risk that substantial value created within our economies is not adequately captured for tax purposes. At the same time, even when new concepts such as significant economic presence are being developed, their effective implementation remains complex, particularly in jurisdictions where the capacity to track, trace and attribute income from cross border and digital transactions is still evolving. Zimbabwe therefore supports ongoing efforts 1 strengthen the definition and application of permanent establishment, 2 advance the operationalization of significant economic presence and 3 ensure that international tax rules better reflect both where economic activity occurs and where services are consumed.

Mr. President, it is equally important that these frameworks remain practical, administrable and responsive to the realities of developing countries. We would be interested to hear from the Panel. How can evolving approaches to permanent establishment and significant economic presence be designed in a way that is both effective in allocating tax rights to source jurisdictions and practical for developing countries to administer particularly given capacity constraints. I thank you.


ECOSOC · President [4:13]:

I would like to thank the distinguished representative of Zimbabwe and I now give the floor to the distinguished delegation of Algeria, Vice President of ecosoc and they will be followed by the distinguished delegation of Burkina Faso.


Algeria · Vice-President ECOSOC [4:30]:

I thank you for covering this important and timely meeting. Algeria aligns itself with the statement delivered earlier by Uruguay on behalf of GSWA 77 and China and by Angola on behalf of the African Group. The current international framework based on the concept of permanent establishment is no longer fully suited to contemporary economic realities. In increasingly digitalized economy, companies can generate substantial revenues in a jurisdiction without having a physical presence there, which limits the ability of the developing countries to effectively exercise their taxon right. In this context, Algeria firmly believes that the concept of significant economic presence represents an important development.

It allows for better alignments of taxation with economic reality by taking into account the revenues generated, interaction with the markets and user participation. For developing countries, and especially for African countries, this approach offers an opportunity to strengthen domestic resources mobilization and better capture the value created within their territory. However, its implementation raises significant challenges. Identifying taxpayers, allocating profits, the issue of double taxation and administrative constraint as well as corporate non compliance remain substantial obstacles, particularly in jurisdictions with limited capacities. Moreover, existing tax treaties may restrict the effective application of these new rules.

Against this backdrop, Algeria would like to make these following points. First, several elements should guide states before implementing the cep. It is necessary to assess the administrative capacities required to enforce these rules, to evaluate the weight of the digital economy in their tax base and to ensure consistency with their international commitments. Second, it is essential to ensure that the measures adopted remain proportionate, predictable and compatible with the objective of economic attractiveness in order to guarantee their acceptability and long term effectiveness. Third, it's essential to adopt a balanced approach.

Significant economic presence should not replace existing mechanisms, but rather complement them in order to address the shortcoming of the current system, particularly in the area of digital services and across border activities. Finally, Algeria, which is actively participating in the ongoing work on the United Nations Framework Convention on Taxation, emphasize the need to achieve inclusive solutions.


ECOSOC · President [7:29]:

I would like to thank the distinguished representative of Algeria. I now give the floor to the distinguished representative of Burkina Faso, to be followed by the distinguished representative of Peru.


Burkina Faso [7:45]:

Thank you, Mr. President. We would like to welcome the initiative of holding this meeting. Burkina Faso subscribes to the declaration made by the G77 in China, as well as the state statement made by the African Group. Ladies and gentlemen, when it comes to modernizing rules to determine taxation links, my delegation considers that the profits made by Big Tech should be taxed equally even when there is no substantial physical presence. Recognizing a significant economic presence is an essential principle to ensure tax justice and ensure the budgetary sovereignty of states.

Burkina Faso already has a tax on added value that is applicable to goods and services sold through e commerce platforms.

This shows our country's will to ensure that taxation is fair when it comes to digital profits and to ensure our fiscal sovereignty. Burkina Faso considers that the value created by users and local markets must be recognized as a legitimate criteria for taxation. The notion of a stable base should evolve to integrate the reality of digital activities far beyond physical frontiers. Significant economic presence must become a general principle which ensures that the profits generated in one country can be taxed equitably. Burkina Faso advocates for an inclusive approach that takes into account the realities of developing countries so that digital taxations can be an instrument for justice and global solidarity.

When it comes to AI, Burkina Faso considers that it represents a major opportunity to transform our public administration. We construct, strengthen transparency and the efficacy of services and ensure a better management of our public resources. AI should serve humans. It should be engineered and used to respect the social principles of equity and digital sovereignty. Its exploitation should be framed by solid legal frameworks in order to ensure the soundness of administrative decisions and preserve citizen trust.

Developing countries must be fully involved in the creation of future future international norms to avoid a greater digital gap. Burkina Faso therefore calls for strengthened international cooperation to share experiences and good practices and ensure that AI can become leveraged for inclusive and sustainable governance. Thank you.


ECOSOC · President [10:43]:

I would like to thank the distinguished representative of Burkina Faso and I now give the floor to the distinguished representative of Peru, to be followed by the distinguished representative of the UN Tax Committee Number four, Madam Kwonju Kiari.


Peru [11:01]:

Thank you, President. Allow me to begin by expressing my recognition to all distinguished panelists for their valuable presentations which have highlighted both the urgency as well as the challenges of modernizing nexus rules applicable to profit derived from from services. Specifically, in the current context of the expansion of the digital economy, this transformative process has been particularly dynamic in developing countries. According to UNCTAD in 2024, services rendered digitally represented approximately 44% of total exports for services in developing economies. In absolute terms, this is $1.1 trillion of value value.

In this context, nexus rules based on physical presence through the concept of permanent base is increasingly less useful to reflect the taxation of income derived from economic activities that are done remotely. Currently, the rules for distribution in treaties to eliminate double taxation usually recognize the right of state state to tax income from services rendered. When this is done through a permanent base in their territory. In practice, these rules limit the ability of many states, including Peru, to adequately tax income derived from digital services and technical assistance, which often is done from abroad. In this framework, Peru considers that the debate on modernizing the nexus rules is fundamental so that the international taxation system effectively addresses the reality of an economy that is increasingly digitalized and service based.

That is why Peru has been actively participating in deliberations in order to prepare a protocol on the taxation of cross border services in the framework of the Interagency Negotiating Committee in charge of drafting the UN Framework Convention on International Tax Quality. In these deliberations, Peru has supported the establishment of nexus criteria that are differentiated based on the kind of service addressing the specific characteristics of these and that the protocol establishes a maximum taxation rate in the source states based on the gross price paid for services, leaving to each state to decide how they will apply the tax on a grosser net basis. The world is changing and therefore the rules that regulate it should change. Peru considers that this current UN process is a historic opportunity to move toward an international tax system that is increasingly inclusive and fair. Thank you very much.


ECOSOC · President [13:53]:

I would like to thank the distinguished representative of Peru and I now give the floor to the distinguished representative of Committee Number Four on Taxation in the United Nations, Wang Shu Kyari, and she will be followed by the distinguished representative of Indonesia.


UN Tax Committee · Member · Wanjiro [14:11]:

Thank you. Chair. My name is Wanjiro. I'm speaking in my personal capacity as a member of the UN Tax Committee.

A comment that was made earlier today is that business practices have been evolving and adapting to globalization and digitalization faster than their rules that govern them. And I agree with that. In relation to the Framework Convention on International Tax Cooperation, one of its objectives is to establish a system of governance capable of responding to existing and future tax related challenges on an ongoing basis. Another comment that was made earlier this week during the Tax Committee discussions is that it's amazing what we can do with AI today that we couldn't do last week. And I think the same thing can be said in relation to global globalization and digitalization that what was being done a hundred years ago when the current rules were developed is not what can be done in the current environment.

So it's clear that there is a critical need for modernization of the current international tax rules. A position that arises continuously during the negotiations of the Framework Convention is that physical presence is still the main factor to be considered in the allocation or attribution of profits or allocation of taxing rights. And developing countries have actually agreed to this but added that it cannot continue to be the only factor to be considered, especially not in light of globalization and digitalization. The Tax Committee has done great work in terms of looking beyond physical presence and will continue to do so in terms of evaluating the PE concept as a whole. The Negotiating Committee of the Framework Convention has been mandated to develop Protocol 1, which is aimed at addressing taxation of income derived from the provision of cross border services in an increasingly digitalized and globalized economy.

The Draft outline also recognizes physical presence as one of the factors to be considered, but the fact that we go beyond physical presence is therefore in line with the mandates that we have been given under the Framework Convention. The Africa Group has also proposed introduction of additional factors such as where the data is being generated from and we know that multinationals are deriving a lot of revenue from the sale of user data without the requirement for physical presence. And I believe that these are factors that are now found in both developed and developing countries. And this is in line with resolution 78 to 30 that says fairness of the international tax rules is in the common interest of all relevant stakeholders. So I urge the Member States and the relevant stakeholders to abide by the mandate that we have been given under the Framework Convention which is in line with the severe commitment that says international tax.


ECOSOC · President [17:13]:

I would like to thank the distinguished representative of Committee Number four on Taxation at the United nations and I now give the floor to the distinguished representative of Indonesia, to be followed by the distinguished representative of Civil Society, United nations association of the usa.


Indonesia [17:31]:

Thank you Chair Indonesia. Welcome this discussion on nexus rules and significant economic press presence in advancing a fair and inclusive international tax system. The SEFIA commitment noted that existing international tax rules do not fully respond to the needs of developing countries, particularly in the digitalized economy. Indonesia views significant economic presence as a complement and in some cases an alternative to the permanent establishment rules.

The SAP better reflects modern business models by recognizing sustainability economic interaction with a market. Even without physical presence, however, trade off must be managed. While significant economic presence strengthen taxing rights for market jurisdiction, it may also introduce complexity, risk of double taxation and fragmentation if implemented inconsistently across jurisdiction. Indonesia support the work to modernize the permanent establishment concept and explore significant economic presence and including the development of clear nexus threshold on implementation. Several challenges remains, including divining clear and workable nexus threshold, attributing income to significant economic presence in a manner that is administrative, administrable and fair, and ensuring consistency with existing tax treaties.

In this context, Indonesia underscores the importance of simplified and practical approaches, including withholding taxes or gross basis taxation. Such approaches can reduce administrative burdens and disputes, particularly for developing countries with limited capacity, while providing greater certainty. Countries should also carefully consider key factors when designing significant economic presence based system, including administrative readiness and institutional capacity, availability of reliable data, compatibility with existing treaty obligation and the need to avoid overlapping taxing rights and dispute. We also wish to underscore that ongoing discussion at the UN must remain inclusive, transparent, development oriented and responsive to the need of developing countries. Indonesia believe that a balanced approach grounded in simplicity, legal certainty, fairness and respect for national circumstances will be essential to ensure that reforms of on permanent establishment, a significant economic presence contribute to a more equitable and effective international tax architecture.

Thank you.


ECOSOC · President [20:02]:

I thank the distinguished representative of Indonesia and I now give the floor to the distinguished representative of the United nations association of the usa, to be followed by the distinguished representative of Belgium.


UNA-USA · Alexa Dominique [20:17]:

Thank you. Chair My name is Alexa Dominique and I'm speaking on behalf of the United nations association of the United States. As a Global Goals Ambassador and also the DMU N Foundation, I would like to commend the panel for addressing the evolving challenges and increasingly digital and globalized economy. Allow me to offer a few recommendations. First, on the role of significant economic presence, SEP may be best understood as a targeted complement to traditional permanent establishment roles rather than a full replacement.

While physical presence provides clarity and administrative ability, it does not fully capture business models where market engagement occurs without tangible footprint. At the same time, expanding nexus through SEP introduces important trade offs. Broader taxing rights may improve alignment with market activity but also increase compliance, complexity and uncertainty. A balanced approach may therefore involve clearly defined their souls, limited scope and predictable application, helping to maintain stability while addressing existing gaps. Second, on implementation and income attribution, one of the central challenges remains determining how income should be attributed in an SEP in a manner that is both practical and consistent.

In this context, countries may wish to consider simplified allocation methods including formula based approaches, safe harbors or withholding mechanisms for certain cross border activities. These tools can support more consistent outcomes while reducing administrative burdens, particularly where detailed information may not be ready available. Ensuring alignment with treaty obligations and minimizing risks of double taxation will also be critical to maintaining cross border certainty. Third, on policy design and country considerations, decision on whether on how to introduce SAP frameworks should take into account administrative capacity, data availability, enforcement, feasibility and potential behavioral responses from taxpayers. Designing rules that are proportionate and administerable will be essential to achieve intended outcomes without creating unintended distortions or compliance challenges.

In closing, modernizing NEXIS rules is a necessary step in adopting to changing business models. The focus moving forward should be on approaches that are predictable, administratable and aligned with real economic activity while maintaining confidence in the international tax system. We thank you.


ECOSOC · President [22:41]:

I would like to thank the distinguished representative of the United nations association of the United States and I now give the floor to the distinguished representative of Belgium, to be followed by the distinguished representative of Cameroon.


Belgium [22:57]:

Excellencies Distinguished Delegates Belgium aligns itself with a statement given by. Cyprus on behalf of the Member States. Of the European Union and complements this with the following national Belgium recognizes that traditional permanent establishment rules may not always reflect the way value can be generated in today's increasingly digitalized and globalized economy, including through remote service delivery and platform based models. Belgium therefore supports continued discussion, including through sharing country experiences, on how nexus rules can be modernized in a way that is fair, practical and consistent with the objective of a stable and predictable international tax framework. At the same time, any approaches should be targeted, administratively feasible and implementable in practice, designed to avoid double taxation and compatible with treaty obligations and with a broad international support.

Approaches should also be compatible with existing treaty obligations and maintain the stability of the international tax framework. It is therefore important to ensure coherence with ongoing work also in other international fora. Thank you.


ECOSOC · President [24:09]:

I would like to thank the distinguished representative of Belgium. I now give the floor to the distinguished representative of Cameroon, to be followed by the distinguished representative of Committee Number six on Taxation at the United nations, man Wanda Montero.


Cameroon [24:29]:

Thank you President for giving me the floor.

We align ourselves with the position of the African group and in this important debate on the definition of significant economic presence we would like to speak about our experience which also is a major challenge for global tax equity. We are developing criteria for measuring the value created by non resident enterprises in our markets and we are observing a reverse and alarming phenomenon that is just a say, the forced elimination of the economic presence of our own economic operators, a marginalization of the global south in monetization mechanisms as well as payments of large platforms, which creates an unprecedented distinction in the notion of permanent establishment.

And. This includes enterprises that have a permanent establishment abroad. Cameroon is developing local financial infrastructure. We have dynamic entrepreneurs that are forced to domicile their activities and their fiscal or tax residence in third party jurisdictions. It is not a choice to engage in tax optimization, but a matter of survival.

Activities that are localized in Cameroon may still lead to permanent establishments that are outsourced in the case of more fluid arrangements. This situation creates a massive erosion of our tax base and we therefore lose the ability to tax not only profits, but also innovation and growth in our own ecosystem. The concept of significant economic presence should not simply be a tool for taxation. It should also be a shield for protecting domestic enterprises. We believe that international tax cooperation encourages digital platforms to interoperate with payment infrastructure that is local.

Without this technical inclusion, the notion of economic presence will remain asymmetrical because it. May.

Continue to encourage the tax exodus or exile of the Global South. The global technical architecture must not prevent taxpayers from existing economically in their own territory. Thank you.


ECOSOC · President [27:24]:

I would like to thank the distinguished representative of Cameroon and I now give the floor to the distinguished representative of Committee Number six on Taxation at the United Nations, Madam Montero, to be followed by the distinguished delegation of Italy.


UN Tax Committee · Member · Montero [27:40]:

Muchas gracias, senor. Thank you very much, Mr. President, distinguished delegations and participants, I would like to to read this statement in my personal capacity to respond to the topic of the prior panel. Countries are facing new business models and structures that are increasingly sophisticated when it comes to fiscal planning. They do not always have the adequate tools to effectively respond.

Even developed countries face difficulties to stay up to date. For developing countries, this challenge is even more serious. This is why capacity development is a central issue. In this sense, the work of the Committee of Experts to which I have the honor to belong, should be highlighted. Their models and guides seek to offer specific tools in order to strengthen the mobilization of resources in developing countries.

Technical progress such as Article 12 AA are important. Nevertheless, their impact will depend not just on the real capacity of countries to apply it, but also on the commitment of developed countries to incorporate these tools in their treaties, new treaties as well as existing treaties. Without this commitment, there is a risk that these steps forward remain only on paper. Our committee is facing a complex task task to perfect the current rules and promote their practical implementation without altering their fundamentals. And here we see an obvious source of tension.

Current nexus rules do not respond to our reality. These rules, moreover, tend to favor the countries of residence versus source countries.

Attribution rules are costly, complex and in many cases not very realistic for developing economies. The lack of data makes analyses highly subjective and demands a high level of specialization and generates frequent controversies. The result is that many countries opt to tax consumption their ec.

They're easy to capture, but they can deepen inequality. So practical guides are valuable, but they are not enough. Without data, without improving the way we provide assistance, and more importantly still, without a deeper reflection on the principles that uphold these rules, the impact of our work would then be limited. As a committee of experts, we have a technical responsibility. But it is countries who should lead the way when it comes to the necessary political decision making to move towards an international taxation system that is more fair, inclusive and aligned with the Sevilla commitment.


ECOSOC · President [30:39]:

Thank you. I would like to thank the distinguished representative of Committee Number six on Taxation in the United nations. And I now give the floor to the distinguished representative of Italy to be followed by the distinguished representative of the Philippines.


Italy [30:56]:

Thank you, Chair. Italy aligns itself with a statement made by Cyprus on behalf of the EU and its member states.

Italy believes that taxing rights in respect of business profits should be allocated to jurisdictions where business act and values created. Therefore, physical presence remains an appropriate basis for establishing tax nexus and allocation rules in most situations. At the same time, we acknowledge the need to evaluate and complement the current standards to respond to the business models in a more digitalized economy. With the view of setting a new forward looking international tax convention, the Intergovernmental Negotiating committee should ensure that the new legal instruments are anchored in future proof principles. The new architecture should aim at establishing a fair and stable economic environment without the distortions that gross based accession may cause and focus instead on net taxation which fosters fairness and equality.

Being aware of the different priorities and the different perspectives, it is essential to work with the aim to reach consensual solutions. We also underline the importance of building on existing frameworks and commitments. The work carried out within this forum must strengthen and complement them by promoting tax cooperation in sectors that are still untapped. Italy remains engaged in a spirit of good faith as negotiations in the international in the Intergovernmental Negotiation Negotiating Committee advance. I thank you.


ECOSOC · President [32:34]:

I thank representative of Italy and now give the floor to the Philippines followed by France.


Philippines [32:47]:

Thank you very much, Mr. President. We align with the statement of G77 in China and thank the panelists. We wish to highlight the following points in our national capacity. The Philippines supports inclusive and effective international tax cooperation and the Sevilla commitment.

We advance a fair, transparent and efficient tax system addressing both domestic and cross border tax leakages. The Philippines engages to strengthen domestic revenue mobilization and overcome barriers to effective tax collection. Our Bureau of Internal Revenue is implementing a multi year digitalization program to enhance efficiency and compliance. And the Philippine Congress has recently passed legislation on taxation of non resident digital service providers. International cooperation remains essential to address cross border tax evasion, base erosion and profit shifting.

Closing tax gaps and developing common solutions to global challenges. Ensuring fair and equitable allocation of taxing rights based on market jurisdiction and value creation is crucial. We should move beyond outdated physical presence concepts and recognize the economic contribution of the market where services are consumed and users located. This includes establishing clear principles that recognize market centric value creation. Fostering flexible and pragmatic solutions remains vital.

Countries should be able to adopt approaches suited to their national context and administrative capacities. Capacity building must be accessible and enhanced and inclusive to promote peer learning and enable effective participation of all stakeholders, especially developing countries. This underscores the urgent and practical need for multilateral solutions facilitated by the UN Framework Convention to achieve widespread and efficient international tax reform. Finally, building on these efforts, greater interoperability and collaboration amongst existing international tax bodies is also essential to leverage comparative advantages and advance inclusive and effective cooperation. We underscore the importance of harnessing digital technologies, including artificial intelligence, to strengthen tax administration while ensuring appropriate safeguards, capacity support and equitable access for developing countries.

I thank you Mr. President.


ECOSOC · President [35:11]:

I thank the representative of the Philippines and now give the floor to France followed by the United Kingdom.


France [35:23]:

Thank you, President. France has long been committed to the principle that business profits should be taxed where value is created. This principle lies at the heart of international taxation and has guided the development of the current system which is based on tax treaties. The existing rules are based on physical presence and the arm's length principle and they remain fully relevant. The criteria for profit attribution based on physical presence through the notion of a permanent establishment or fixed place of business, supplemented by transfer pricing principles are well adapted, relevant and appropriate.

You can see this through the near universal adoption of the arm's length principle. Although it is not a minimum standard, these criteria are reliable, robust and broadly understood by tax administrations and by taxpayers. They must remain the only international frame of reference for the allocation of taxing rights between jurisdictions. However, in light of the major economic transformations of the past few decades, certain business models are not are poorly captured in these principles. That is why France for over a decade now has advocated for adapting these rules in accordance with sectors and situations when the disconnect between physical presence and the creation of value is too great.

Work is already being undertaken in this regard. It must now be completed. However, any broader change of international tax rules would be ill advised because first of all, it could reduce the importance of physical presence, which remains a cornerstone of the international tax framework and has been for decades. Secondly, it could also cause taxation on a gross basis that is too far removed from economic realities. Lastly, France would like to recall the major progress achieved in international taxation over the past few decades.

These advances must be protected and strengthened by the actors that initiated them and that have specialized expertise. The work in this area must not undermine these achievements. On the contrary, it should strengthen and complement the MTAS cooperation in still underdeveloped or emerging areas such as environmental taxation. Thank you.


ECOSOC · President [37:31]:

I thank the distinguished delegator of France and I now give the floor floor to the United Kingdom followed by the International Chamber of Commerce.


United Kingdom of Great Britain and Northern Ireland [37:42]:

Thank you Chair, panelists and delegates for this discussion. The UK welcomes the opportunity to contribute. To this special meeting and support the. Continued constructive engagement on international tax cooperation.

The UK's long standing view is that taxing rights in respect of business profits should be allocated to to jurisdictions where businesses act and create value. In line with inclusive framework discussions on taxation of the digital economy, the UK considers that there is a need to. Evaluate whether delivering this principle in a more digitalised economy requires modernisation of existing. Nexus and profit allocation rules. At the same time, we remain of.

The view that physical presence should remain the foundation of nexus and profit allocation. Rules and continue to provide an appropriate. Basis for establishing tax nexus in most situations. Should jurisdictions decide to reconsider nexus and. Profit allocation rules, the UK considers it important that these issues are considered together with the profit allocation being the more important and challenging consideration.

The UK also has significant concerns about. Any greater reliance on gross based taxation as part of any changes to taxing rights. Giving the Given the potential for economic distortions and barriers to cross border trade, we look forward to continuing discussions with. A shared objective of maintaining a stable and efficient international tax system. Thank you.


ECOSOC · President [39:15]:

Thank the distinguished delegate of the United Kingdom and now give the floor to the International Chamber of Commerce.


ICC [39:23]:

Thank you Mr. President. Thank you very much to the panelists as well for sharing their views and country experience as icc. We would like to briefly share three short points. First, while we understand there might be tax challenges emerging from the digital economy, current permanent establishment rules remain highly relevant. Brick and mortar business do still exist Even though digital driven business models have increased.

The vast majority of business models around the world are still very much anchored in physical presence and operations. Current PE rules are still very much of central importance and broader challenges can lead to an incredible level of uncertainty for taxpayers and tax administration alike. Second, a reference was made to the burden of VAT that differently from other tax measures or approaches will fall on consumers. We would like to highlight that when it comes to gross basis taxes as well as DSTs, these are and are already being passed on to consumers. On top of their potential detrimental effects on trade investment as well as ensuring broad access to digital connectivity, we highly recommend taking into account the broader economic effects that these provisions have and as a Starting point we would refer back in relation to Article 12 AA to the independent report conducted by Oxford Economics.

As ICC we keep favoring a non tax based approach as well as supporting the need for avoiding double taxation over taxation. Finally, stable, clear and predictable rules are fundamental for businesses around the world, globally, regionally and domestically, especially in today's geopolitically unstable environment. We recommend strong coordination among efforts undertaken within the UN both at committee of experts and INC level as well as with other international tax fora that are working on the topics discussed today in gispane. Thank you very much.


ECOSOC · President [41:12]:

I would like to thank the distinguished representative of the International Chamber of Commerce. We have heard the last speaker for this segment. I now give the floor to Mr. Mongibola so he may invite the panelists to provide their concluding comments and also for a closing statement. Thank you.


Moderator · Mr. Mongibola [41:32]:

Thank you very much Mr. President. And so I would turn to Mr. Katishomo. There was a question from the floor which was how SEP can be designed for simplicity and practicality. If you will please kindly respond to this question as you also include your closing statement in three minutes. Thank you.


Panelist · Mr. Kajeshoma [41:59]:

Thank you. Mr. Mondreto.

It is important that when you are designing an SCP it's important that we undertake planning and as I said earlier there are three tripod of taxation, the policy, legislation and administration. So it's important that you design how the SCP will work around these three tripod. So another thing you need to think about is to make it simple, it should be complementary because of course it's possible to have an SCP that act as alternative because it depends on how you define the SAP. But that will be extremely complex. So to make it simple it is better to focus on a SEP rule that is complementary.

Then in the legislation you may also adopt the double legislative process that is you have the primary legislation and you have the secondary legislation. The reason is that in most jurisdictions primary legislation are very difficult to change. So so have the main rules included in the primary legislation. Then have your secondary legislation to specify the detailed rule and in case there are issues you can easily amend those issues in your secondary legislation. Then it's good to have a SEP rule that allows for self assessment returns.

That is you allow taxpayer to file their returns and you can take it off from there then administratively it's important to deploy technology you can it's difficult to tax digital economy with a tax administration that is not digital. So as you are looking towards implementing SEP then you also need to be thinking of digitalizing the tax administration processes, capacity building is also very important. You also need to look at that then profit allocation. One of the initial challenges Nigeria had was when we developed our SEP rule we didn't include profit allocation. And the outcome we have businesses turning 0.01 or 0.001 of their turnover as profits or as the tax.

So it's important as we design the SEP rule you include profit allocation because the transfer pricing rule will not work for SEP because if you remember price transfer pricing rule allocates profit based on function, risk and asset and this will not be present in the source jurisdiction. If you are looking at SCP then you need to look at strong political will to punish defaulters. Because if you are unable to punish the photos, most businesses will not comply. So for those that are complying voluntarily they need to be encouraged and those that are not complying you must be able to for instance you may be should be able to block their Internet from assessing your jurisdictions. Then we may also need to consider administrative efficiency.

One of the canons of taxation is economy. So administrative efficiency in terms of low compliance cost for the tax administration and also low administrative cost for the tax administration, low compliance cost for the businesses then collection how do you intend to collect? So it's important that you have a simplified registration and compliance process. Because if most of these business are not your jurisdiction and you don't expect them or most of them will not want to come to your judicial to file returns or to to comply with the processes. So you should be able to create a process where from their jurisdiction they can register, they can file returns, they can make their payments.

So that is also very very key. And lastly because of so that I don't exceed my time lastly SCP will work better in multilateral context and that is why we are discussing here and discussion is going on at different fora including at the inc. So it is important because the current rule and of course that's one of the challenge. India also mentioned it the fact that even when you have the SCP it does not override your tax treaty. Because the taxation for under tax treaty is based on SCP.

I mean based on Article 5 permanent establishment rule. So it is important we have a multilateral context that's able to bring on board this new rule so that is incorporated into the treaty models and treaty rules and everybody is able to implement similar rule in different jurisdictions. Thank you.


Moderator · Mr. Mongibola [47:30]:

Thank you very much Mr. Kajeshoma. Ms. Kenner, what do you see as likely trade off between a unilateral domestic rule and a broader international tax Framework acceptable to many.

Please. As you look at that question, you may also want to make your concluding remarks. Three minutes. Thank you.


Panelist · Ms. Kenner [47:59]:

Thank you very much, Mr. Moderator.

I think.

The experience for small economic, small countries, small economic countries is that, you know, you really have to work together in order to be able to have a consensus based solution to all your problems. It's, you can be very right, very correct and do anything correct. But if you are a small actor in the big world, you do need a multilateral solution, especially on the cross border issues. As I had mentioned in my country, we have a solution to some, you know, using the indirect taxes, but as pointed out, you know, in the end of the day that is also costly to the consumer. So although it doesn't require an international agreement and therefore I perhaps think, think there are better solutions.

But there are solutions. So, you know, the better solution is to do something multilateral. And I find it very positive to hear the consensus in this room supporting the work of the international tax system. And perhaps as a just small, short technical comment, I understand that there were quite a few delegations that advised that they do not agree on a gross taxation, but they are actually not against taxation and changing the nexus rules. And I find that so.

I understand the comment that we can change the nexus rules, we can change this balance of who has the right to tax as long as it's made on a net basis. And I find that very encouraging. In the work that we have done so far, we have identified gross taxation with a net taxation as an option for the taxpayer. And that seems to be, for the moment at least, a very good way forward. So thank you for all your comments and it's been very interesting to listen to you.


Moderator · Mr. Mongibola [50:16]:

Thank you. Thank you very much, Madam. Professor Alfredo, from a legal perspective, what you think is the most appropriate approach to analyze SEP in relation to conventional Lexus rules? And as you please look at that question, you may also want to make your concluding remarks. Thank you.


Panelist · Professor Alfredo [50:48]:

Thank you very much, Mr. Moderator. I would like to offer a series of concluding reflections and remarks and among them I'll ask. I'll answer the question that you just asked. First, obviously I would like to thank the distinguished representatives for their statements and interventions. They were very interesting and they highlight the clarity within this debate and also the need to finally carry out this task as a real priority for the international tax cooperation community, which is to adjust nexus rules.

And second, and this specifically is to answer your question, I think that beyond the SEP rule, there's a number of alternatives that have been highlighted they are different in scope and they can complement existing nexus rules. And I think that complement could be a basis at least from a legal standpoint. And what we see at least hearing the statements from today is that a one size fits all approach is probably not an adequate solution to find international consensus around this issue. I think that if we want to move forward and I think we should be able to, it is necessary to work jointly and to find points of convergence and to build bridges of understanding amongst the different visions that exist in terms of this necessary update for, for the nexus rules in order to tax current business profits.

And from there I think it would be a good idea to develop a plan that is progressive and flexible.

I don't think it would make a lot of sense to try to solve problems that are impossible from the very get go. And finding those points of convergence should start from the most simple issues and move toward the more complex ones. And we mentioned some alternatives that I think could be a good starting point for us. And for that I also think that it would be important to bear in mind that ultimately what's at stake is not just equity amongst nations and the adequate assignment of tax competency among countries, but also competitive neutrality among different economic operators and ultimately tax equity amongst taxpayers. What many of the participants here highlighted to or they spoke to a fair, robust and inclusive system, which is what we seek.


Moderator · Mr. Mongibola [53:49]:

Thank you. Thank you very much. Professor Alfredo, I will turn to you. Respondent. Mr. Chaudhry, you've listened to the floor, you've listened to the closing comments of the panelists.

Maybe in two minutes. Can you just give us your reflection again?


Respondent · Mr. Chaudhry [54:11]:

Thank you. Thank you Mr. Moderator. And we appreciate very much the practical feedback given by the panelists on how to actually design significant economic presence taxes and how to implement them. And we again reiterate that developing countries need not wait till a multilateral solution comes out because they can immediately go ahead and start implementing these taxes. And today's discussion provides valuable experience which.

Can be built upon to introduce such taxes. And these can also feed in to. International standards both within the UN Tax. Committee and within the UN Framework Convention. It is also very clear from the discussion that markets users is the basis, a fair basis for the allocation of.

Taxing rights and nexus. And this should be also reflected in. International standards both in the UN Model. Convention and within the Service Protocol. Thank you.


Moderator · Mr. Mongibola [55:10]:

Thank you very much. Mr. Chaudhary. I think I'll sum up this way. It's very clear from the comments and the presentations of the panelists, the various Comments from the floor and the summation of the respondent that the current definition of tax measures strictly on the basis of physical presence is no longer adequate for for the modern day business. And also there are various options that could be considered.

Some are already in practice in different countries and there are some that could come and solve specific problem. It's also clear that whatever is going to come to complement the current definition of tax measures must be acceptable to all so that it can become practicable other than having unilateral action and also not to over tax businesses. I know that all of these are very important issues that the task committee will work with in moving forward is working in this membership and we also believe that will also be considered by the members delegates who are negotiating at the inc. But one thing is clear.

The world cannot remain the same the way it was a hundred years ago. With this I want to say very thank you to my panelists and also to all respondents. Chair and Mr. President, thank you very much for the opportunity. Opportunity.


ECOSOC · President [57:00]:

I would like to thank Mr. Buongibola for the expert win which he has managed our debate as well as the panelists for the contributions and delegations for their participation in a productive exchange of points of view. Distinguished delegations. I will now briefly pause the meeting to allow a change in the podium for our third panel discussion. Please remain seated.


Speaker 42 [57:46]:

It.

I'm going to a different meeting, but after that.


ECOSOC · President [1:01:17]:

Good afternoon. We now move on to the third panel discussion entitled Artificial Intelligence and the Future of Tax Administration.

This panel will explore how AI can strengthen revenue collection and compliance while ensuring appropriate safeguards for taxpayer rights and public trust. I am pleased to welcome the distinguished presenters for this discussion And I invite Ms. Geneva Jacqueline Bangura, Commissioner General of the National Revenue Authority of Sierra Leone and member of the Committee of Experts on International Cooperation in Tax Matters to introduce the panelists and proceed with the discussion.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:02:18]:

Thank you very much, Mr. President. I am truly honored to moderate the final panel of this special meeting which will focus on the use of artificial intelligence by tax administrations. AI is transforming how tax administrations operate worldwide. AI can improve efficiency, detect fraud and evasion, improve risk assessment and strengthen taxpayer services and compliance management. Yet AI also introduces significant risk, increasing privacy concerns, data vulnerabilities, ethical considerations and potential errors that could adversely affect both tax administrations and taxpayers rights.

Drawing on experiences from tax administration at the different stages of AI adoption, this panel will examine how to harness AI's potential responsibly and inclusively. Identifying governance standards, capacity building priorities and cooperation mechanism that can help all countries benefit from technological transformation while promoting tax rights and maintaining public trust. I will engage this panel to address the following questions. 1. How can AI enhance fairness, efficiency and transparency in tax administration and what are the key risk to guard against?

2. What are the main opportunities and key challenges countries face when integrating AI into tax administration, particularly developing countries? 3. What skills, infrastructure and institutional arrangements do tax administrations need to deploy AI affiliate efficiently and responsibly in areas such as risk compliance, risk management, risk assessment and audit selection and finally, what governance standards and safeguards are needed to ensure the ethical use of AI in tax operations, including human oversight requirements, requirements, transparency and disclosure, data confidentiality protections, sound algorithm design and accessible redress mechanisms for taxpayers. Last but not least, before introducing our distinguished panelists, I would like to highlight the work being undertaken by the UN Committee of Experts on International Cooperation in Tax Matters in this area.

The Committee is currently developing a practical guide on Artificial Intelligence for tax Administration which is expected in 2027, designed with particular focus on developing countries. I am confident that this guide will benefit from from the country experiences and priorities shared during this panel. Let me now introduce our distinguished panelists who are representing governments, international organizations and academia. To my Left we have Mr. Juliano Brito Digusta Neves and is Under Secretary of Corporate Management at the Federal Revenue Authority of Brazil. Online with us we have Mrs. Zapor Alexandia.

She is the Deputy Head of Tax Information and Risk Analysis Department at the State Revenue Committee of the Report Republic of Armenia. Also online we have Mr. James Smitten, who is also who is the Director General of Enterprise Transformation and Chief I A Chief AI Officer at His Majesty Revenue Majesty's Revenue and Customs in the UK to my far right I have Mr. Richard Stern, who is the Director of the Global Tax Policy center at the Institute for Austrian and International Tax Law, Vienna University of Economics and Business. Finally, as a Panelist we have Mr. Abubakar Emir, who is the Head of Facilitation Information Systems and Strategy at the Moroccan Tax Administration. We also have two distinguished respondents in the audience with us this afternoon we have Mr. Rud and he is the Deputy Director of the Fiscal Affairs Department at the International Monetary Fund. And lastly we have Ms. Angela Eng and she is the Assistant Commissioner at the International Tax and Relations Department of the Inland Revenue Authority of Singapore and she also doubles as the member of the UN Committee of Experts on International Cooperation in Tax Matters.

Now, before turning our first to our first speaker, let me briefly just outline the format of this session. Each panelist will have not more than five minutes to deliver his or her initial remarks. I will then turn to our two respondents for intervention for their reaction, and then we will come back. I will turn it over to the president, who will invite interventions from the floor. I therefore now encourage all panelists to be concise, candid and concrete in your intervention and to focus on the key messages and actionable recommendation.

We will start with Mr. Stern, and I would like to ask if. I would. If you would please start by giving us a general overview of the current state of AI adoption by tax administrations around the globe, as well as the main lessons you have learned from the research you have conducted in these areas. Thank you.


Vienna University of Economics and Business · Director · Richard Stern [1:09:48]:

Thank you, Madam Commissioner. Thank you, Mr. President. And thank you to the Echo Soc for having me here and to my distinguished panelists and. And participants. Five years ago, if we started talking about AI, you might have looked around and wondered, what are we talking about?

It sounds like this magic bullet that is going to solve all of our problems. I'm guessing at this point everybody has. Has dabbled in. In AI and generative AI, maybe in Copilot, and can see what the power is, even at a personal level. So what I'm going to do here is to give you an overview of how AI is used in tax administrations.

Talk about its power, which has a positive and negative outcome. What I want to start with, though, is three important foundations for you to keep in mind for the panel and beyond. The first is AI is a tool. It's a means and not an end. And that's really important because when people who don't haven't used AI say, I'm going to use AI.

Now, the question is, should be in your head for, for what? A second foundation is that AI is powerful because it's a tool and it goes alongside a function. So not all AI is alike. It's not all used the same way. And so when we talk about AI, it's important to distinguish what it is we're using it for.

And finally, for a foundation, it's important to know that AI doesn't act alone. AI is driven by a model, and AI is powered by data. And having those two components behind it is what makes AI so powerful. And also, that's where the risk comes from. And it's important to keep these in mind.

So what do tax administrations use AI for? Well, it uses AI in lots of different ways. And I think all around the world you'll see that just about every administration is using some form of AI. I like to think of AI in Two different sorts of buckets. One is used for internal purposes, so to help processes become faster, more efficient, and without mistake.

That's AI used for things like writing letters, doing research, pulling together documents, maybe even doing a first pass of tax returns so that it can flag very obvious errors. AI on the internal side is also used for simulations, doing models. If there's for example, a spike in oil prices, how will that affect my tax compliance and how will it affect my tax take? AI can help you with all that. And then there's AI, which I call used for external purposes.

And that's AI used directly with taxpayers. And in that respect, it can be used for things like taxpayer service, making taxpayer service available 24 hours a day with efficient chatbots that can not only help you with a problem, but, but can also give you the documentation you need and guide you through it to assessing risk and flagging a taxpayer that it determines through the model that is not operating within the bounds of the tax code. It can also be used, and perhaps this is one of its most powerful attributes, to pull data together from different sources. It used to be that when tax administrations wanted to look at third party data, so social media data or purchase data of advertisements about what you're buying these days, you would do that by hand. Now, with the help of an AI tool, it can be done in nanoseconds, pulling data in from all different parts of a taxpayer's activities and strengthens the picture that you build.

When you look at taxpayer data, that is extremely powerful and that is also the source of a lot of controversy. So when I think of tax, the use of AI in tax administration internally versus externally, what I'm really doing is probably proxying for risk. When tax administrations use AI for internal processes, the level of risk, if things go wrong are very low. When it uses AI for external processes that affect a taxpayer, then the risk of things going wrong is high and the consequences are high. So one of the things that we've done over the past couple years is to collect cases, cases where AI has gone wrong.

And we've learned a lot from that. And I want to share that with you for the remainder of my time. AI, when it goes right, can do, can really transform tax administrations. For example, there's a small country in Europe that has used AI for 10 years, and through AI, it's raised its, its tax take by something like 4, 40% off of audits, meaning it gets the audits right more often than not. So it, it is more productive.

It's used something like 50% fewer resources to do the audits because its AI tool is picking the right people and it's, it's lowered processing time by 2/3. All of that is good for tax for taxpayers and building trust and tax certainty. Now let me turn to a couple of examples of when AI went wrong. I think many of you know about a famous case that happened in Europe three years ago involving childcare benefits where the model was wrong, the model was biased and chose the wrong people to pull out for, to flag for risk, for high risk, and then stopping the services by something like 96%. So we learned a lot from that.

And I'll come back to that in a second. There were a couple other very high, high profile cases which showed that the data that wasn't being used was, or the data that was being used was wrong. And in a third case, everything was right except that the tax administration didn't have the legal foundation to actually access this data and use it. So when it was challenged, the court ruled in favor of the taxpayer and the tax administration had to stop. So what have we learned from some of these bad cases?

We've learned a lot. First and foremost, we, we've learned that there has to be a person who oversees any AI enhanced tool or process. Without an adult in the room, things are bound to go wrong. And all of the big profile cases happened in countries that you would never expect that to happen in because there was no one checking the model. The second thing is the model itself.

Models have to be checked regularly to make sure that they're biased. And equally as important, they give you the results that you're looking for. It's not that they give you results, they give you the right results. So they need to be checked. Third, you need a legal foundation that allows you not only to use the model, but also to access and use the data.

And that's the data part of this, that when we talk about some of the guardrails that might be built to help countries as we build them for the use of AI, we also need to make sure that they're there for data and there needs to be transparency. And the last thing I want to leave you with is that there's a lot of work being done about AI and taxpayer rights. And the challenge there, and hopefully we'll come back to this in the discussion, is that most of the legislation that would protect a taxpayer doesn't exist in the tax administration. It exists in other parts of the law. And that makes it hard to control tax administrations.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:19:02]:

So thank you very much thank you. Very much, Mr. Stern. Clearly, based on your intervention, the AI has changed significantly over the last five years. And of course we will expect for that to continue as more countries adopt the use of AI in tax administration. So thank you very much.

Over to you please. Mr. Hemere, can you please share with us the benefits but also the risk that your government has encountered while implementing the use of the AI and your work in the Moroccan tax administration. Thank you.


Moroccan Tax Administration · Head, Facilitation Information Systems & Strategy · Abubakar Emir [1:19:46]:

Thank you. I will speak in French.

So first I would like to say that it's a real pleasure for me to participate in this panel and share the experiences of Morocco's time tax administration when it comes to using AI tools. First, I would like to say that I completely agree with what my dear neighbor just said in his statement and I do also agree with all of the statements we heard in prior panels. When we speak about AI, the first thing that comes to mind is that this is something that we use on a daily basis. This is part of life every day now. It provides us with many services and usually we use it without thinking too much about it, especially our children.

They were born into this face recognition, the Internet of things, among others. And it's true that more and more tax administrations are, are using AI based tools because as you said, they strengthen equity effectiveness, compliance and they also help us fight against fraud. But one thing is certain, their deployment has to be done in an organized fashion. It needs to be logical. What this means is that in short, this is a gradual progress process.

We go by phases and this happens throughout a spatial timeline. It's not just a single event. So it's important to really regulate these tools to make the most out of all of them. And in Morocco, we started by digitalizing our activities and processes so that all tax matters can be handled online. We provide taxpayers with all the relevant information online.

There is no need for the taxpayer to go to a physical office. Now this digitalization also means the availability of an important mass of data internal extract online. And for us, it's very difficult to make the most out of this data because we don't necessarily have tools to deal with all of this macro data. And as you repeated and recalled, due to issues that have to do with security, confidentiality, updating the data, accessing the data, and this is why we started using AI based tools. And on the one hand, they helped us to gain experience.

This is important. And when we talk about experience, we're thinking of economic forecasting, planning, programming, budgeting perspective and predictive analysis. We're thinking of evaluating the impact of a decision or tax provision being able to follow the money. But it's much more, more than that, the idea is to try to detect the gap between a goal and a result and to adjust accordingly based on the relevant parameters. That's one aspect.

And another aspect is that these tools have allowed us to improve our operational piloting because now we can cross reference all kinds of data based on, on a certain number of rules and analysis and very specific indicators so we can declare any kind of, so we can detect any kind of anomaly, we can determine whether there's any kind of irregularity in the numbers. And that way we can identify at risk task payers. And this is on a surface level, the use of data and the use of analytics that is helping us improve our decisional intelligence. And this we complemented by using algorithms that we call intelligent. We speak about deep learning or machine learning programs that can learn and that allow us simply to create models for situations and behaviors, to create profiles.

And through that we are able to, to identify at risk taxpayers and guide our actions. And beyond that, to fight against fraud more effectively, perhaps not eradicate it, but make it harder for people to commit fraud. So from all of this there are a series of consequences and benefits for the Moroccan tax administration. And let's be honest about this for us, but also for taxpayers. We can talk about improving productivity, about optimizing our treatment of the data because we're automating many tasks, because we have higher quality interventions and decision making, because we are able to better target fraud cases, and we have the capacity to provide personalized services through chatbots, through virtual assistants, and, and at the end of the day, the idea is that only taxpayers that need specific personalized support or help with complex issues come to our physical tax administration.

And so the idea is having an impact based management.

We're not going to be trying to improve compliance, compliance, etc. We're tax behaviors. That's all good. But there are requirements here, requirements for us considering data as a strategic asset that we ought to make the most of. When I say strategic asset, I mean the fuel.

It's the fuel behind algorithms and models. And we need to have robust and resilient infrastructure equipped with appropriate tools for treating this volume of data. And that's when you look at our experience. And we know that there are a number of pitfalls, threats, anticipated, risks. When we talk about our experience, we know that we have to avoid over the dimensional projects that are ambitious, costly, with sometimes useless investments that are abandoned along the way.

We need to Ensure that we have tools with clear benefits in the short term already. And then all of that brings us back to the question of cost. Yes, we do need to look at infrastructure, material, equipment that may be extremely costly. We also need to think about dependency, dependency on editors.

And there might be solutions that do not meet our needs. And for many of us, that includes open source tools. We have the problem of the cloud, sovereign cloud. That's the question of where are we going to store our data outside the national territory. And so we have data centers at the national level.

The other topic to talk about is data security. When it comes to cyber threats, many of us will have specialized units for cybersecurity in the government. And then there's the question of human resources, that is the people who are responsible for doing the modeling, modeling situations, behavior and translating rules into algorithms. So how do we keep and train these personnel? And then lastly, regulation.

For most of us, this regulation was developed well in advance of the Internet and AI rules on protection of data privacy, confidentiality, cyber threats. We need all of those regulations. If we deal with the issue of governance, that's how we get to a responsible and ethical use of AI tools. So, so that is the general approach, the approach, the consequences, the requirements and the risk that we've anticipated and that we have experienced from trying to share with you. In summary, thank you, thank you very.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:28:38]:

Much Mr. Emir, for sharing your government's experience. It appears that your way advance the adoption of AI using tax administration. And what I clearly got from your submission is that before even you consider using AI, digitalization is very important because you need data. Data is what you need to use AI. And then you also mentioned that you know, during implementation you should be ready and available to adjust to your work plan as you move along.

And of course having the right human resources, right infrastructure, you need to have the right trained human resources to ensure that the use of AI in tax administration is seamless. Thank you very much for sharing your experience. I will now welcome Mrs. Alexander. She's online. And Mrs. Is she, is she on?

Please welcome. And we know Armenia has been leading pilot projects using AI to improve the efficiency of tax administration. Could you please tell us a bit more about these projects and also how Armenia manages the inherent risk posed by the use of AI. Over to you, ma'. Am.


State Revenue Committee of Armenia · Deputy Head, Tax Information & Risk Analysis · Zapor Alexandia [1:30:05]:

Dear colleagues, it's a great honor to contribute to this discussion on the role of AI in tax administration. AI is rapidly reshaping tax systems worldwide, offering new opportunities to strengthen domestic resource mobilization, improve compliance and enhance efficiency at the Same time, it raises important policy questions related to to fairness, transparency, taxpayers rights and governance. In this context, discussions under the United nations, including in follow up to the civilian commitment, are essential to ensure that AI contributes to more inclusive and effective tax systems rather than widening existing gaps between countries. The risk of an emerging AI divide is is particularly acute for developing countries facing constraints in data infrastructure, institutional capacity and access to advanced technologies. Against this backdrop, Armenia's experience reflects a gradual and pragmatic approach to integrating data analytics and emerging AI tools into tax risk management while ensuring appropriate safeguards and institutional readiness.

In modern tax administration, data analytics plays a central role in managing compliance risks and improving interactions between tax authorities and taxpayers. Many processes are now conducted through automated non contact systems that identify inconsistence and notify taxpayers electronically. This enables taxpayers to correct declarations voluntarily, supporting cooperative compliance while allowing tax authorities to focus enforcement efforts more efficiently. In this regard, digital tools including AI should be seen not only as strengthening as enforcement but also as enhancing trust and transparency. AI further improves risk assessment by enabling the analysis of large datasets to detect patterns and anomalies, allowing authorities to better target high risk cases while reducing unnecessary interventions.

At the same time, its use requires careful attention to data quality, cybersecurity and algorithmic bias. Ensuring that AI systems are explainable, auditable and subject to human oversight is therefore essential to safeguard taxpayer rights and maintain public confidence. At the same time, many countries, especially developing economics, face significant challenges in in integrating AI into tax administration, including limited infrastructure, data quality issues, skills shortages and fragmented data systems. Addressing these challenges requires not only technological investment but also capacity building, strong data governance and enhanced international cooperation. The State Revenue Committee in Armenia currently operates more than 200 rule based analytical models which serve as a foundation for identifying compliance risks and ensuring data reliability before introducing AI tools.

Building on this foundation, we have gradually introduced AI and machine learning pillar models in selected sectors, focusing on targeted applications that allow us to test methodologies and build institutional capacity. One example relates to compressed natural gas transaction where inconsistencies arise due to different measurement units and variations in product descriptions. Using AI based text analysis, we are able to identify relevant transactions transactions across large data sets and detect discrepancies between reported quantities and turnover. Another example focuses on dairy production chains where AI helps compare data reported by producers and buyers to identify potential undeclared turnover while overcoming challenges related to INC consistent product descriptions. These examples illustrate how AI can address common challenges related to unstructured data and fragmented reporting systems.

At the same time, effective deployment of AI requires strong analytical infrastructure and high quality data. It's also essential to ensure government protection frameworks, human oversight and safeguards for data protection and transparency, including a clear accountability mechanisms and accessible redress systems for taxpayers.

Ensuring that developing countries, including landlocked and structurally constrained economies, have access to the necessary tools, infrastructures and expertise will be essential for achieving more equitable outcomes in domestic resource mobilization. AI offers significant opportunities to strengthen tax compliance systems, but its success ultimately depends on responsible governance and strong institutional capacity. If deployed in exclusively and responsibly, AI can contribute to more efficient, fair and transparent tax systems supporting sustainable development and the achievement of the Sustainable Development Goals. Thank you very much.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:36:32]:

Thank you very much, Mrs. Alexander, for that intervention. Mania seems to be doing a lot. What I gathered from your submission, you know, in the implementation of AI and use of tax administration, it's, you know, it helps enhance trust and transparency between the tax administration and then the taxpayer, but most importantly, the general public as well. So thank you so much for sharing that experience with us, Julian. Now you're welcome.


Federal Revenue Authority of Brazil · Under Secretary Corporate Management · Juliano Brito Digusta Neves [1:37:10]:

And I know that Brazil has been doing a whole lot, so please just share with us the Brazilian Revenue Authority's experience in implementing AI, in particular, with regards to the governance standards you have set for the ethical use of AI in tax administration. Thank you. Thank you, Madam Commissioner. It's a pleasure to be here. And we have seen the dawn of lots of international guidance and directives regarding how to handle AI, how to have the ethical use of AI.

But I would like to rephrase that question. How can we implement those directives in a real world scenario back in Brazil? The Federal Revenue of Brazil has just recently published its new AI policy based on more than 15 years of experience using AI way before the generative AI put the subject into the spotlight. And what are the most important things of this new policy that makes AI ethically sound in Brazil? First and foremost is that you have to have human oversight and autonomy.

That's clear. The most basic of all directions. You. You need human control as a basic principle of AI. But not only human oversight and human control.

AI should not condition or bind the exercise of legal competence. It's written right there in our policy. The human agents owns the final decisions and is the sole responsible person for this decision. And we think that's so important that we already will have put in our own policy that if you do not follow the AI guidance and AI policy, the officer is personally can be charged with personal offense, disciplinary offense. In the Brazilian disciplinary model, inside the revenue Administration, AI must be.

The AI policy must be followed. Or we say that our AI policy has thief because people should know that they must not step outside of the line when using AI. And when doing that, we also created a new role regarding how to handle AI. We create the generative AI curator, a special officer that has the responsibility to oversee everything a generative AI does. Monitoring, hallucination, bias and accuracy.

And also has a kill switch. If the AI is going rogue, he it can stop the AI before it can go further and create lots of problems.

We have said here, we have seen here saying about sovereignty of data in Brazil. Since the beginning of 2023, we started doing a lot of bringing our data together and create data sovereignty as a country. We have a big discussion with all the big tech solutions and how to bring their data centers to Brazil. But not only that we use all those software and the hybrid inside the Brazilian government companies handling AI. And also that the big tech cannot access our data.

We have air gapped our system and we are doing the same thing with AI. And that's a lot of difficulties because usually the special hardware needed for running AI is only on the US or Europe. But we must have the big techs going to Brazil. So we put that in our policy. It's a governmental directorate and we are challenging them to do so.

If they are not willing to do to bring the machinery inside our own premises, we are not going to do business with them. And also we put on our policy that we will not hire or procure any kind of solution when the third party vendor does not sign that they will not use our data for training their own model. If they cannot guarantee that we are not going to hire them, procure or source anything from them. Any AI solution in the Brazilian Revenue Administration must underwent at least three assessments of risk. First one inside the business unit, then the IT department should see the initiative and made a risk assessment.

And last but not least, we have an IT committee When all the business units have a seat on and they we we discuss together as an organization. If that AI solution is following our risk tolerance declaration. We have business units that are more tech savvy and wants to use AI for everything. We have others that are more conservative and you must have a plain analysis of risk throughout the organization. So we put that in our policy.

But it is interesting to see that if you have a very strong risk assessment, you you may make it difficult to innovate. So we create a fast track for innovation. If you are doing AI on prototypes, experimental solutions and they are on a controlled environment, separated from our Production one, using only synthetic data, you can go on without the risk analysis. Afterwards, you must go through all the risk analysis before the solution is put in place, put in production. Last but not least, we created a permanent training plan for our staff.

People should know the probabilistic nature of the technology of AI and not believing that with your whole heart. You have to know that there are limits to that. And when building our AI solutions, we try always to use those that are transparent. That's possible to explain what is doing. We know that's impossible to audit what a large language model is doing.

But we should be prepared to explain to the taxpayer how the AI using its data, remembering then that the final solution is always, always, always a human one. So in a NutShell, that's our AI policy. We handle lots of things that we think that is an interesting way to implement all the international directives. And we are more than willing to share our policy with any country that wants to know more. Dear Madam, thank you so much.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:43:40]:

Thank you very much, Liano, for sharing the Brazilian experience. Clearly, the policy puts the taxpayer in the center of everything that you do. And clearly you have established that AI, the use of AI in tax administration is just an enabler. You still have the human factor need to make sure that decisions are still made by tax administrators that's wholly and solely dependent on AI. So thank you very much for sharing that experience.

So I'm going to move back with James online, hmrc. James, we know that the UK has been leading AI adoption in government. Could you please share your experience in particular with the card.

With regard to the potential of AI for taxation, tax administration, but also the key enablers in terms of skills, infrastructure and institutional arrangements for a successful AI deployment. Thank you.


HMRC · DG Enterprise Transformation & Chief AI Officer · James Smitten [1:44:54]:

Thank you very much. Thank you, Madam speaker and everyone. Thank you for the invitation and it's great to be here to represent HMRC in the uk. Yes, I think definitely many of the answers to that question you've already heard from some of the panel. It is really important in order for us to have an AI system that is fair for everybody and effective to have the basics in place.

So there are three things that, that we have prioritised in the UK and in hmrc, and the first one has already been touched on and that is data. None of the rest of this works without investing really effectively in data. And I don't mean just in the creation of solutions like a data lakehouse, which we are currently in the process of completing, the finishing of, but also in the way that we build very carefully the ingestion of data into our organization, the way that we enable the safe access of the data in our organization and the way that we agree carefully the arrangements in which we can share data across government. As a tax authority, we sit on very, very valuable data that is incredibly helpful for the smooth running of many government services. And as the speakers have already said, we therefore have a huge burden upon us to responsibly look after that data on behalf of customers.

And so probably the most important bit for us is that investment in the way that we store and the way that we access our data and the way that we are able to share it with other services. The second piece obviously is technology and the ability to to have been able to invest in cloud based solution software as a service technologies. And so we have been able to make investments and are in the middle of procuring new foundations for HMRC in the UK that will allow us to provide a single joined up view of the customer and allow us to significantly leverage AI out of the box. So whilst we have many, many really effective AI solutions, some of which we have built ourselves, our main strategy has been to make the most of some of the really interesting and new pioneering solutions that are available out of the box today in CRMs and in other contact center solution products. So technology is the second critical one, but the third one I would really call out, which we've touched on today, is the human and probably the hardest part of AI transformation and exploitation in any tax authority in the world today.

It is about how we invest in our people and how we invest in creating the right culture, in hiring people with that deep curiosity and that tenacity to be able to help us really drive innovation in AI. How we make sure that the organization does not get lazy by the use of AI, but in fact remains robustly and strongly the human in the loop. And as speakers have already touched on, some of the surprising incidents that have occurred around the world around AI have often come from humans, not from the AI itself. And so that is probably the biggest challenge for us is how do we make ethics and how do we make the statement human in the loop actually practical and workable, so that we have processes that are well controlled and well overseen and well monitored and ultimately we keep the human beings as the accountable people running the system and do not hand it over to AI. So in the UK we have been able to invest in one of the largest rollouts in UK government of Microsoft Copilot already touched on and that is already having a very significant Impact.

We are over 30,000 users now in the UK. We are shortly about to switch on the agent capability and have had to invest significantly in training and support to help colleagues have the confidence to use it and to assure ourselves that the data is protected, that it is safe and that the way we use it is responsible so that it genuinely drives productivity in the organization and ultimately delivers better services for customers and helps us to collect tax. The use that we've had up until now in general, AI and advanced analytics has enabled us to save reduce the tax gap by over 8 billion British sterling in 2024. 5. So we're already seeing very material benefits as a result of the use of AI and we can see that only increasing.

I think the big challenge for many tax authorities is keeping up with the incredible strides in AI capability and making sure that we walk before we run and that we have really good and strong disciplines in place to make sure that we responsibly use it. Because if we do not do that, we will not build trust and if we do not build trust, we will not have active tax authorities.

I hope that is helpful. I'm very happy to talk about any of the specific cases that we have deployed in the uk. Thank you. Thank you very much, James, for sharing the UK experience. Clearly you have established that, you know, an effective implementation or deployment of AI revolves around making sure that you invest in your people, invest in the institution, because you need to make sure that you have the right staff, trust, trained, need data.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:51:04]:

Without it, I mean, you don't want, you know, as Juliano said during his policy, their policy, people centered, right? You protect the taxpayers, but you ensure that the staff that continues to work with the taxpayers are trained. So thank you so much for sharing the UK experience to us. And with that, I just want to say thank you to all of our panelists for their rich insights. And I will now turn over to our two respondents in the audience.

I'm going to start with Ms. Ange. She's again the Assistant Commissioner at the International Tax and Relations Division of the Inland Revenue Authority of Singapore and she also serves as a member of the UN Committee of Experts on International Cooperation and Tax Matters to react to, you know, what our panelists have submitted and we would be particularly interested in hearing on how these relates to your experience in Singapore. Over to you and thank you.


Inland Revenue Authority of Singapore · Assistant Commissioner · Angela Eng [1:52:23]:

Good evening and good afternoon, everyone and thank you to the President and moderator for this opportunity to respond to a very impressive panel. Our panel has spoke about the AI's transformative promise. I think Mr. Stern talked a lot about the power of AI, the lessons learned from the good and the bad. And I think we have heard today, from today's conversations repeated calls for, for robust governance frameworks to deal with the risks and capacity building efforts to ensure that developing countries can harness AI benefits. And so I believe the imperative to start on this AI journey is clear.

For those starting out on the AI journey, my takeaway actually, especially hearing from the other tax administrators, is to pilot with clearly defined use cases. In Singapore's instance, we started actually with just three data analysts building a predictive AI model using open source tools that did not require extensive coding to improve audit case selection. And over a few years we saw recoveries coming in three times higher than any other by random sampling. So for those who are already progressing on AI, our call is that our strategies need to be dynamic and responsive because AI is evolving at such unprecedented speed. And so for this, Singapore likes to have frameworks.

And so I'd like to share with this committee a framework. We have this framework that puts people human at the center. And that's what we hear from all the speakers so far. And there are four E's to this. The first is educate to constantly upskill and foster and AI for everyone mindset.

So AI doesn't mean no humans, but humans continue to have oversight and autonomy and they need to be trained for that to happen. The second E is empower to establish governance structures from the start. I think we hear that from the Moroccan delegate and also from Brazil. Similar to Brazil, we adopt a risk based approach. We assess every use case based on the level of AI autonomy and the extent of potential harm.

And low risk ones can be automated readily while high risk ones need explainability testing and human oversight. The third E is experiment pilot use case. I think the piloting mindset is very clear in the Armenian panelists. Our moderator herself also shared that as a UN Tax committee, we are embarking on a practical guide on AI and focusing on use cases. And so my fourth E is enterprification.

I think we made up that word. Essentially it means pilot, pilot, pilot. But let's scale up, let's scale up and let's make it embedded in our processes and systems. And I think that was a point that Richard brought up, that in the end it's not AI on its own. It has to serve a function and it has to build in our functions.

We have heard again and again that AI is a tool, it is an enabler. And so I want to go with the saying that says many Hands make light work and AI is newest and most possibly powerful hand that we have been given. So let us as humans use these hands practically responsibility, responsibly, but we can do good things with these new hands we have. Thank you.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [1:55:58]:

Thank you very much for that. We'll now move on to our second respondent, Mr. Dumoy, and he is the Deputy Director of the Fiscal Affairs Department at the imf. Please share with us. In addition to your initial reaction, perhaps we would also like to hear main takeaways from the recent work the IMF has done in this area. Thank you very much, Very much, Chair.


IMF · Deputy Director · Mr. Dumoy [1:56:28]:

And thank you very much to the panelists for a very insightful discussion. So let me respond on what we've heard so far and also illustrate that this is all very consistent what we see in the work that we do in our member countries in technical assistance. So an important message I guess that I take away is AI is revolutionizing revenue administrations, not only tax administrations, also customs. I think that's important to emphasize, very important for developing countries and especially because these are very data driven processes. And as we've just heard, it can potentially have major gains in the effectiveness and efficiency of revenue administrations.

And here we've done earlier research on digitalization. We've heard how important digitalization is ultimately for reaping the gains from AI. If digitalization is done well, we find that this can create significant gains in terms of revenue in the order of 0.5 to 0.7% of G. So these are major implications for domestic revenue mobilization. But we've also heard that there are risks and concerns. And Rich had a very interesting example from actually the country where I'm from, I'm from the Netherlands and it illustrates the dramatic impact that it could have if you use biased data in your algorithms.

This led to a huge scandal which ultimately led to the resignation of the government. So this is how biased data through AI in a revenue administration can have large consequences. And from all this also lessons from other countries. I think we have learned important lessons and we have heard them today from various panelists. The the role of humans is so important, but it's just not enough to say, well, we should just keep humans in the loop.

We should operationalize this and how to make this work. I thought it was very interesting to hear James say humans should not get lazy because of the AI. I thought that was a very good one to keep in mind. We should really improve the competencies of people. That's the essence.

Training new skills. But more generally, I think the Lesson is also that AI is not adopting a new technology. It's a comprehensive transformation of the whole organization of processes, of people and of governance. So I think these are important lessons, but these are lessons drawn from, I think, revenue administrations that are relatively sophisticated. And I think it's important to also note that, that there is still a huge digital divide in the world.

I very much liked the phrase that was said by Morocco. Before you get to AI, you need digitalization, you need the data. And because there is this digital divide, there is also an AI divide in the world, which is important.

So many low income countries, many small states do not have the advancements in the new technology yet. So what is the holdup there? Well, part of it is resources, of course. So there needs to be some investment, not just the investment in the technology, but also investment in the people and in the organization in the transformation. There's also legitimate concerns as the risks that have materialized have shown.

Biased data, confidentiality issues, transparency, black box kind of algorithms that are very difficult to explain. So how do you deal with this? How do you get to a responsible and effective use of AI? And this is where we as the imf, as providers of technical assistance can help. We help with our capacity building programs to navigate this process for countries.

We organize peer learning events where we bring countries together. Next month during the spring meetings, we have an event especially on this for revenue administrations. And we have a series of publications that highlight how you can do this, practical guides on how you can adopt AI responsibly in your organization. And one specific one is on generative AI. Now one of them has a 10.10 step towards responsible AI.

And I'm not going to go through all 10 steps, but the first step is really interesting. That's exactly what Brazilian experience illustrates. Have a AI use policy which is actually not quite common in countries, but an AI use policy is really the best practice. You should specify the goals, ethical principles, the legal aspects, transparency policy, the use of external AI services, governance by humans, managing risk training. So that's the number one in the list.

After that comes sort of dealing with all the use cases, as we've heard, existing use cases, maybe also emerging use cases. What are the use cases relevant for you? Because countries differ. It's not one model for all countries differ. So this is kind of the basis, maybe.

I have two small final remarks. AI is revolutionizing revenue administrations, but it's also revolutionizing tax analysis and tax forecasting. Countries are using the wealth of data, for instance on VAT and payroll data. For nowcasting which can help you improve, forecasting, which can help improve your policies ultimately. And second is that advances in administration can ultimately also drive advances in policy.

Things we could not do before, we may be able to do now or in the future with AI. And I think Brazil again is a good example with its view VAT reform, where it introduces a progressive VAT which utilizes a lot of the new information for the design of the policy. But there's many other examples you can think of, maybe accrual based capital gains taxation to all other taxes. So let me stop here.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [2:03:20]:

Thank you. Thank you very much to our two distinguished respondents for your sound reflection. Truly appreciate it. You've captured all that we have discussed, the panelists have discussed. So I want to say thank you to you both for that.

I now turn the floor back to Mr. President. Thank you.


ECOSOC · President [2:03:47]:

I would like to thank Madam Bangura for moderating our discussion and I would like to thank the discussants for their substantial contributions to our debate. I now invite delegations to engage with the presenters in an interactive discussion. Those who would like to request for the floor should press their microphone button. I would like to remind speakers that in order to allow for a large number of delegations to speak, time limits for statements on behalf of groups will be set at five minutes and three minutes for individual statements.

I now give the floor to the distinguished representative of the Russian Federation, to be followed by the distinguished representative of Sweden.


Russian Federation [2:04:35]:

President the Russian Federation has importance to the introduction of modern digital solutions, including artificial intelligence, the technology technologies with a view toward enhancing the efficiency, transparency and fairness of tax administration. The Federal Tax Service of Russia applies AI to technologies to improve the quality and reliability of state registries of legal entities. The approach is based on the use of self learning predictive models that generate risk profiles for taxpayers through analyzing various types of data. And the system produces a ranked list of legal entities with a high probability of inaccurate data and this allows for the automated prioritization of audit activities. In 2025, the use of AI made it possible to identify hundreds of thousands of cases of inaccurate reporting, contributing to the updating of the State Register of legal entities and enhancing the transparency of the business environment.

AI technologies are also used in the Russian Federation to improve mechanisms to manage tax arrears. These solutions are implemented through a specialized analytical platform that allows for the integration and processing updated from multiple sources. The system allows for a step by step analysis including identifying taxpayers with risk indicators, an assessment of their activities in depth, as well as comparison with relevant judicial practice, and preparation of analytical materials. With recommendations on further measures. These solutions contribute to improved detection of complex tax evasion schemes, reduces the time required for analysis and ensures greater consistency in decision making.

Thus, the experienced version Federation demonstrates the introduction of AI technologies subject to the appropriate safeguards can serve as an effective tool for modernizing tax administration. And the Russian Federation is prepared to share of course, its experience in this area. Thank you.


ECOSOC · President [2:06:48]:

I would like to thank the distinguished representative of the Russian Federation. And now I give the floor to the distinguished representative of SWEET Sweden, to be followed by the distinguished representative of the U.N. association of the U.S. thank you.


Sweden [2:07:02]:

Chair, Excellencies, distinguished panelists, delegates across the world, tax administrations are indeed under growing pressure to mobilize domestic resources more effectively and in an increasingly complex and digitalized environment. For many countries, this challenge is compounded by limited access to technology, data and infrastructure. Therefore, this discussion is very helpful for sharing experiences, best practices and lessons learned.

Digitalization is an important part of Sweden's international development, cooperation and support to long term capacity development. This is evident in the extensive peer to peer support provided by the Swedish tax agency that is working directly with partner tax administrations globally. Our starting point is always to understand the specific, specific demands and problems that need to be addressed. Digital solutions are often, but not always part of the answer. For many partner countries, the path forward will require addressing both access to technology and access to experience.

Artificial intelligence is often presented as a powerful opportunity. Our experience suggests that while technology is necessary, it's never sufficient on its own. The decisive factor is how the transaction transition is organized and led. In light of this, allow me to highlight the three main takeaways from the experience of the Swedish tax agency. First, identify the concrete needs for the production already from the outset.

Even where advanced technology is available, AI delivers value only when it becomes part of regular operations. Second, invest in the right combinations of skills. AI in public administration requires close collaboration between technology domain expertise and an understanding of legal and ethical responsibility. Without this institutional capacity, technology alone cannot deliver results. And third, establish governance.

Early responsible use of AI depends on clear roles, transparency, human oversight and alignment with legal frameworks. This is essential to protect taxpayers rights and maintain trust in public institutions. For Sweden, it was important to start by creating a small dedicated team with a clear mandate supported by the senior leadership that provided direction and resources. The aim was to deliver real value and avoid piling up pilots that never reached production. The team was deliberately cross functional, combining technical expertise with deep knowledge of tax processes, data and legal context.

This made it possible to work across organizational boundaries and develop solutions through continuous learning and gradual scaling. The approach produced steady progress, tangible results and a growing set of AI supported services in daily use. This also built trust and legitimacy, enabling leadership to sustain and expand the effort. We look forward to continuing to build on our partnership and collaboration in this important area. Thank you.


ECOSOC · President [2:09:59]:

I would like to thank the distinguished representative of Sweden. I now give the floor to the distinguished representative of the UN association of the US to be followed by the distinguished representative of Indonesia. Thank you.


UNA-USA · UNA-USA · Alexa Dominique [2:10:13]:

Chair My name is Alexa Dominique. Speaking on behalf of the United nations association of the United States as the Global Goals Ambassador and also for Diem Yun Foundation, I would like to commend the panel for addressing a critical critical dimension of modern revenue administration and rapid technological transformation.


UNA-USA · Alexa Dominique [2:10:29]:

Allow me to make a few recommendations. First, on the role of AI in tax administration. AI presents a significant opportunity to enhance compliance management, improve risk targeting and optimize resource allocation. When effectively deployed, it can reduce administrative costs, strengthen revenue performance and improve taxpayer services through more timely and accurate interactions. In the system sense, AI has the potential to support more efficient and responsive tax systems.

Second, on implementation challenges and disparities the benefits of AI are not evenly distributed. Their differences in digital infrastructure, data quality and institutional capacity can affect the pace and effectiveness of adopt adoption. As countries consider integrating AI tools, it may be important to prioritize scalable and modular approaches, allowing administrations to adopt solutions aligned with their operational readiness while avoiding complexity or cost burdens. Third, on governance and safeguards, the effectiveness of AI will depend not only on technical capability but also on trust. Clear governance frameworks covering data protection, transparency, human oversight and accountability are essential to ensure that automated systems support fair and consistent outcomes.

Poorly designed systems or insufficient safeguards may introduce risks that could undermine confidence and over time affect compliance behavior. Fourth, or international cooperation. Enhanced coordination and knowledge sharing and technical support can help reduce fragmentation and support more consistent implementation. Practical guidance developed through UN processes can help can play an important role in scaling workable solutions and improving administrative efficiency across jurisdictions. Enclosing the integration of AI into tax administration offers clear opportunities to improve performance and service delivery.

Realizing these benefits will require approaches that are proportionate, well governed and aligned with institutional capacity while maintaining confidence in the integrity of tax systems. We thank you.


ECOSOC · President [2:12:31]:

I would like to thank the distinguished representative of the UN association of the US I now give the floor to the distinguished representative of Zimbabwe, to be followed by the distinguished representative of indonesia.


Zimbabwe [2:12:48]:

Thank you, Mr. President. We thank the panelists for sharing their practical experiences in the use of AI in tax administration. In particular, the insights of what has worked and what hasn't have been very helpful. AI presents significant opportunities to strengthen tax administration, particularly in areas which include risk analysis, compliance monitoring and audit efficiency.

Zimbabwe recognizes this potential and has taken steps towards the automation of tax administration, including digital filing systems and improved data management processes. This approach is further guided by the country's National Artificial intelligence strategy of 2026-2030, which positions AI as a driver of inclusive development, economic transformation and national sovereignty. Importantly, the strategy emphasizes that countries such as Zimbabwe must not remain passive consumers of important technologies, but active developers of context driven solutions solutions. However, we believe it is important to situate the discussion on AI within a broader context. For many developing countries, the challenge is not only about adopting advanced technologies, but about ensuring that the foundational systems required to support them are in place.

These include reliable digital infrastructure, high quality data systems and adequately resourced tax administrations. Without these foundations, there is a risk that the rapid advancement of AI could widen existing gaps between tax administrations rather than help close them. Zimbabwe therefore supports a progressive and inclusive approach to AI in tax administration. One that aligns with national capacities, is supported by targeted capacity building. It is complemented by investments in corporate systems and infrastructure.

Ultimately, AI should be viewed as an enabler of stronger tech systems, not a substitute for building them. I thank you Mr. President,.


ECOSOC · President [2:15:00]:

I thank the distinguished representative of Zimbabwe and I now give the floor to the distinguished representative of Indonesia, to be followed by the distinguished representative of Algeria, Vice President of ecosoc. Thank you Chair. AI offers significant potential to enhance tax administration, including through improved compliance, risk management, anomaly detection and more targeted audits. In Indonesia, these efforts are being advanced through the development of the CORTEX system as part of broader tax administration reform. This represents one of Indonesia's key efforts to strengthen domestic reform.


Indonesia [2:15:37]:

Resource Mobilization in line with SEVILLA commitment, CORTEX integrates end to end tax processes and centralized taxpayers taxpayers data, enabling more effective use of analytics and automation. AI is utilized to support knowledge management, data collections and analysis, helping in identify risks, improve compliance strategies and support both extensification and intensification of revenue collection. This reflects Indonesia's broader approach to leveraging digitalization to enhance efficiency, transparency and evidence based policymaking. And based on our experience, allow me to highlight four key considerations. First, the benefits of AI must be balanced with appropriate safeguards.

Risks related to bias, lack of transparency and misuse of taxpayer data must be carefully managed to maintain trust, accountability and fairness. Second, effective implementation requires reliable data, robust digital infrastructure and sustained investment. And in this regard, Indonesia underscores the importance of ongoing work on the tax administration and digitalization, including emerging discussion on AI, as well as strengthen international cooperation, capacity building and technical assistance. Third, institutional readiness is essential. This includes skilled human resources, secure and interoperable system and clear operational frameworks to integrate AI into tax administration processes.

Technological innovation must be aligned with national priorities and capacities. Fourth, governance frameworks are critical. AI must ensure human oversight, protect confidentiality and uphold transparency and accountability. IT support, not replace human decision making, particularly in sensitive areas such as audits and dispute resolutions. Chair Indonesia stand ready to share experiences and contribute to international cooperation to ensure that the US of AI in tax administration is inclusive, responsible and development oriented, benefiting all countries.

Thank you.


ECOSOC · President [2:17:47]:

I would like to thank the distinguished representative of Indonesia. I now give the floor to the distinguished representative of Algeria, Vice President ecosoc, to be followed by the distinguished representative of Mexico.


Algeria · Vice-President ECOSOC [2:18:01]:

Thank you, Mr. President. AI is profoundly transforming text administrations around the world, as highlighted by the Sevilla commitment. It represents a major opportunity to modernize domestic resource mobilization, improve administration efficiency and recent tax compliance.

With AI, Tax administration administrations can more. Effectively detect fraud and tax evasion, refine. Risk assessment and provide faster, more tailored services to taxpayers. When used responsibly, it can help make. Tax systems more equitable, transparent and efficient.

However, this progress also comes with significant risks. Concerns related to data protection, lack of transparency and automated errors must not be underestimated. Without adequate safeguards, the use of AI. Could undermine taxpayers rights and erode public trust in tax institutions. Moreover, we observe a widening gap between countries.

Some already have advanced system, while others, particularly the developing countries, still lack the necessary infrastructure, technical expertise and appropriate governance frameworks. Without collective action, AI risk exacerbating these inequalities. In light of these challenges, Algeria believes. That several priorities are essential. First, capacity building is crucial.

Tax administration must invest in skills, digital. Infrastructure and data quality. Second, it's vital to establish robust governance frameworks. This includes human oversight of automated decisions, system transparency, strict data protection and accessible. Appeal mechanism for taxpayers.

Third, Algeria believes that international cooperation is. Crucial to advancing the modernization of domestic. Resources, mobilization and improving administrative efficiency. In this context, developing countries must have access to the technologies, expertise and the. Resources needed to fully benefit from AI.

Finally, Algeria encourages the ongoing work of. The Committee, particularly through the development of. A practical guide on AI in the administrations, which represents a highly relevant initiative. For the developing counties. I thank you, Mr. President.


ECOSOC · President [2:20:32]:

I would like to thank the distinguished representative of Algeria. I now give the floor to the distinguished representative of Mexico, to be followed by the distinguished representative of Belgium.


Mexico [2:20:46]:

Mr. President, Mexico would like to thank you for organizing this dialogue as well as the panelists for their valuable statements recognize the contributions of Enrique Volado, especially when strengthening international tax cooperation. In the context of Sevilla, AI is transforming tax administration by automating processes, detecting patterns, and supporting decision making. Nevertheless, its adoption in the public sector is conditioned by the need to protect rights, ensure equity, and comply with strict transparency and accountability standards. From Mexico's perspective, AI should be understood as part of a system that integrates technology, data capacities and normative frameworks. This is why the challenge is not only to adapt these tools, but doing it in a way that strengthens the legitimacy of the tax system and creates value for the public in terms of opportunities.

This technology can contribute to greater equity by making the identification of risk more aware and reducing discretionality. It can also improve efficiency through the automation of tasks and strengthen data analysis, improving thus the allocation of resources. However, its use can also carry significant risk. Opaque or poorly designed systems can increase bias, create errors, and affect the privacy of taxpayers. This underlines the importance of strengthening transparency, explainability and the possibility to question automatic decisions.

For developing countries, AI is an opportunity to strengthen the mobilization of internal resources and modernize their tax administration. Nevertheless, these opportunities are conditioned by limitations when it comes to technical capacities, data quality, infrastructure, and normative frameworks. Overcoming these challenges requires sustained investments in human capital, specifically in data science and algorithmic governance. It also demands integrated information systems and adequate management and data protection systems. In this context, Mexico considers that it is essential to strengthen institutional capacities.

This includes multidisciplinary teams, monitoring tools and models that are traceable. Moreover, it is necessary to establish robust governance frameworks based on a risk focus that ensures significant human supervision, transparency in the use of automated systems, and protection for fiscal data. In this sense, international cooperation is key in order to share standard strength and capacities and avoid new asymmetries when resources are mobilized. Mexico would like to reiterate its willingness to continue collaborating with Member States and other relevant actors in order to promote the responsible, inclusive and equitable use of AI in tax administration in order to contribute to sustainable development. Thank you very much.


Belgium [2:23:26]:

I would like to thank the distinguished representative of Mexico and I now give the floor to the distinguished representative of belgium. Thank you, Mr. President, and thank you. For organizing this very insightful and inspiring panel and the valuable examples also provided by the respondents. Belgium fully supports modernizing revenue administration, including responsible use of AI and advanced analytics, detecting non compliance and addressing revenue leakage,. Thereby strengthening compliance services and domestic revenue mobilization, which is consistent with Sevilla's focus on modernization and better revenue data.

Belgium emphasizes this responsible use of AI in Tax administration to ensure taxpayer rights, confidentiality and data safeguards. Thank you.


ECOSOC · President [2:24:20]:

I would like to thank the distinguished representative of Belgium. I now give the floor to Madam Bangura in order to invite panelists to provide brief final responses and to make her own closing remarks.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [2:24:37]:

Thank you very much, Mr. President. And I would like to also join you to say thank you to all the contributions from the distinguished members. State.

What I take from the interventions from countries is AI is not just a simple solution. What I hear from the delegates is technology alone cannot achieve results. Digital projects fail when they are not embedded in a wider context. Zimbabwe mentioned also the broader context, and also it runs the risk of achieving the opposite if not done well. So in terms of creating greater trust, certainty and tax morale.

So I think this is a very important lesson. One final remark, which I found very interesting, is the intervention from Morocco. We're only at the beginning, and I think the technology itself is evolving very rapidly. And if we think about what will happen in the future, I think we have to. I mean, we are all talking about developing models, but in a few years from now, AI will develop models for us.

Thank you very much. In closing, James, over to you, please, online.


HMRC · DG Enterprise Transformation & Chief AI Officer · James Smitten [2:25:59]:

Thank you very much. I mean, I think I would say AI doesn't succeed in a tax authority because it's powerful. It succeeds if it's trusted. We must not outsource our brains. This is about keeping the human at the center of everything that we try to do.

In harnessing it, we must use it to solve real problems. I think the biggest challenge that I'm hearing is a thousand flowers blooming. The danger of doing lots of little small experiments to look good so that people feel like they. They've joined the AI club rather than actually using it intelligently, at scale to solve the real problems that they have. And I think a final thought.

I'm really excited about how the technology is developing, and I think it makes it possible for more tax authorities to be able to transform their services because it's more affordable and more flexible than the technology that we've had in the past. So it's exciting to see how this is going to continue to disrupt the tax authority. And I think if we can do it responsibly, it will be amazing to see what we can do across the globe. Thank you very much.


Sierra Leone · Moderator · Geneva Jacqueline Bangura [2:27:19]:

And with that, I would like to thank our panelists, respondents and all participants.

Today's discussion has only underscore that AI is not simply a technological shift. It is a strategic opportunity to strengthen fairness, efficiency and trust in Tax systems worldwide. I hope the insights shared here will inspire continued collaboration and thoughtful action. I wish all delegates safe travels. And now hand over the floor back to Mr. President.

Thank you.


ECOSOC · President [2:28:03]:

I'd like to thank you, Ms. Mangora, for your leadership of our discussions, which are very interesting. And thank you to all of the panelists for their very important contributions. Thank you to delegations as well for taking part in these very productive exchanges. Ladies and gentlemen, I suggest that we take a brief break now so we can rearrange the podium for the closing session.


Speaker 80 [2:28:46]:

Sa.


ECOSOC · President [2:31:14]:

Excellencies, distinguished delegates, we have reached the conclusion concluding segment of this meeting. I now invite Ms. Bijok Sankhyar, Assistant Secretary General for Policy Coordination in the Department of Economic and Social affairs, to deliver closing remarks.


DESA · ASG Policy Coordination · Bijok Sankhyar [2:31:41]:

Thank you very much to the ECOSOC President and Excellencies, distinguished delegates, speakers, ladies and gentlemen, as we conclude today's special meeting, I would like to express my sincere gratitude for the richness of today's discussions and for your valuable contributions. I think the exchange today reflected the breadth of perspectives from member states as well as diverse stakeholders, really underscoring how multilateral dialogue shapes more inclusive, fair and effective tax systems that support sustainable development. We've heard how discussions also highlighted how closely international tax cooperation is linked to the broader financing for development agenda.

As reflected in the Sevilla commitment. It was very clear that domestic revenue mobilization is essential to sustainable development and that tax policy must align with development objectives. Panelists and member states offered practical perspectives on how to advance both at national and international levels. The Panel on Permanent Establishment and Significant Economic Progress Presence illustrated concretely how international tax cooperation must evolve in response to a fast changing economic environment. As more economic activity takes place remotely and across multiple jurisdictions, physical presence alone no longer adequately captures where economic participation occurs.

Now, while some states have adopted national policies to this new reality, we heard how discussions highlighted the importance of continued cooperation to address the challenges of taxing cross border activity in a digital economy. We also discussed the growing role of artificial intelligence in tax administration.

As we heard in the discussions just before the closing, tax administrations shared firsthand experience using AI to prioritize audits, to identify fraud to facilitate compliance in both developed and developing countries. Several speakers also raised a question that I think is worth carrying forward. What can AI enabled tools do to improve the core quality of tax policy making itself? From analysis and forecasting to assessing the distributional impact of policy options. The panelists emphasized the importance of effective safeguards and human oversight, particularly for the most sensitive decisions.

They stressed that international cooperation and capacity building will be essential to make these tools accessible to countries at all levels of capacity. Throughout the day, Member States and stakeholders underscored the important role of the United nations, in particular the Committee of Experts on International Cooperation in Tax Matters, in providing guidance and tools to navigate a rapidly evolving tax landscape. Discussions on strengthening international tax cooperation will also continue in negotiations towards a United Nations Framework Convention on International Tax Cooperation, and the reflections shared today may contribute to these ongoing deliberations. Ladies and gentlemen, please allow me also to acknowledge our international partners, including those participating in the Platform for Collaboration on Tax, several of whom were represented in today's panels. The messages from the PCT Tax and Development Conference in Tokyo earlier this month on the urgency of domestic revenue mobilization, on translating country ownership from principle to practice, and on making digital tools accessible to countries at all levels of capacity found strong echo in today's discussions and added force to them.

Fair and effective taxation at both national and international level is central to financing for sustainable development. The special meeting gives a clear path forward. Progress requires continued cooperation, practical solutions and sustained commitment. And ladies and gentlemen, please be assured that the United nations remains committed to supporting Member States in this effort. Thank you.


ECOSOC · President [2:36:40]:

I thank the Assistant Secretary General for Policy Coordinations in the Department of Economic and Social affairs for her closing remarks. Let me now deliver my own closing remarks. Excellencies, distinguished delegates, participants, as we come to the conclusion of the ECOSOCUS Special Meeting on International Cooperation in Tax Matters, allow me to thank all the participants for your valuable contributions throughout the day. Today's discussions have highlighted the central rule that international tax could coordinations cooperation plays in supporting countries efforts to mobilize domestic resources and achieve sustainable development. Our first panel examined developments following the SEBIYA commitment and the importance of strengthening domestic resource mobilizations.

We also heard insightful discussions on permanent establishment and significant economic presence, highlighting the challenges countries face in adapting tax frameworks to an increasingly digitalized and globalized economy. Participants shared valuable national experiences and perspectives on the opportunities and challenges associated with the modernizing NEXUS rules. Finally, our discussion addressed the growing role of the artificial intelligence in tax administrations. These exchanges highlighted both opportunities offered by new technologies and the importance of appropriate safeguards and governance frameworks to ensure that they are used responsibly. Excellencies, distinguished delegates, Today's meeting has demonstrated the value of bringing together policymakers, experts and stakeholders to exchange experience and perspectives on emerging issues in international taxation.

Allow me once again to thank all the speakers, panelists and the participants for their engagement. Strengthening international cooperations on tax matters will remain essential as countries continue their efforts to mobilize resources for sustainable development. I thank you all for your participation. Excellencies, distinguished delegates, we have thus concluded our program of work for this afternoon. Before closing, I wish to remind delegation that the first special meeting of the Council credit rating is mandated by the provisions of the Sebiyya commitment will be held next Monday, the 30th of March, beginning at 10am in this chamber.

The meeting is adjourned.