Developments in the internet governance environment in July to September 2025

Wolfgang Kleinwächter Oct 15, 2025 28 min read

In the third quarter of 2025, the discussion on Internet governance was dominated by the following five topics:

Preparation for the WSIS+20 review conference
Artificial intelligence, the role of the UN and the positioning of the major powers
Progress of UN negotiations on cyber security
Discussion on state oversight of critical internet resources
New negotiations on a global digital tax

I. WSIS+20 Review Conference

Preparations for the WSIS+20 Review Conference entered the critical phase in the third quarter of 2025. Following a further round of consultations on the Elements Paper published in June 2025 [1], the two co-facilitators from Albania and Kenya published a zero draft [2] of the planned final document on 29 August 2025. The zero draft contains 148 paragraphs divided into 18 chapters.

a. In terms of content, the Zero Draft incorporates criticisms expressed during the consultations on the "Elements Paper". The Zero Draft makes positive proposals for continuing the WSIS process in the next decade. The document is based on the WSIS documents from Geneva and Tunis (2002–2005), such as the commitment to a "people-centred, inclusive and development-oriented information society". The paper calls for overcoming the digital divide and supporting the multi-stakeholder approach to Internet governance. Against the backdrop of geostrategic tensions, the document offers positive compromise formulations for controversial issues (enhanced cooperation). The Zero Draft documents progress in the implementation of the eleven WSIS action lines, identifies existing deficits (infrastructure, financing), responds to new technological developments (AI, quantum), seeks close coordination with the implementation of the UN Sustainable Development Goals (SDGs) and the Global Digital Compact (GDC), and outlines a perspective for the next ten years (WSIS+30). The moderate tone of the Zero Draft is both a strength and a weakness. On the one hand, it aims to avoid destructive discussions and a possible failure of the WSIS+20 negotiations. On the other hand, the vague generality of many recommendations leads to a certain irrelevance.

With regard to the digital divide, the draft notes that the number of Internet users has risen from 15% in 2005 to 67% in 2025, but that two billion people still do not have Internet access. Today, the digital divide is primarily manifested in excessive costs for Internet access (affordability) and a lack of training (digital skills gap). A lack of financial resources is the main obstacle to the development of an information infrastructure in the Global South. AI could further widen the divide (AI gap). Moderate criticism is levelled at concentration processes in the global digital economy, which lead to unfair competition and discrimination. Greater investment in smaller companies, especially in developing countries, is called for. The issue of a global digital tax is not addressed.
In the chapter on cybersecurity, the zero draft remains very general. Security and trust in cyberspace are prerequisites for innovation and sustainable development. Cybersecurity must not be played off against human rights. Criticism is levelled at "all forms of hate speech and discrimination, misinformation and disinformation, cyberbullying and child exploitation and abuse". There are calls for "capacity building" and "awareness raising". The multi-stakeholder approach should also be strengthened for cybersecurity. There is no reference to the various UN negotiations on cybersecurity (OEWG, GGE LAWS, UN Cybercrime Convention).
The chapter on human rights refers to the "universality, indivisibility, interdependence and interrelation of all human rights and fundamental freedoms," which apply both offline and online. It calls for "an open, safe, secure, stable, free, interoperable, inclusive, accessible and peaceful digital technology environment." Human rights must be taken into account from the outset when developing technology. Companies are reminded of their obligations under the UN Guiding Principles on Business and Human Rights.
On the subject of data governance, the Zero Draft limits itself to references to the GDC and the work begun in May 2025 by the UNCSTD Working Group on Data Governance (WGDG).
The chapter on artificial intelligence is structured in a similar way. Here, too, the Zero Draft refrains from making its own recommendations and refers to the GDC and the proposals contained therein for the establishment of an "International Scientific Panel on AI" and the launch of a "Global Dialogue AI Governance". Particular emphasis is placed on the need "to increase AI research expertise in the Global South". The UN Secretary-General is called upon to set up an "AI capacity building fellowship" for developing countries.
The chapter on internet governance begins with the definition of internet governance from the Tunis Agenda. Since 2005, this definition has been considered the most important international reference for the multi-stakeholder approach.
1. In contrast to the Elements Paper, the Zero Draft emphasises that multilateral and multistakeholder processes must go hand in hand. It points out that Internet governance encompasses both political and technical processes (encompasses both technical and public policy issues and should involve all stakeholders). The "open, free, global interoperable, reliable and secure nature of the Internet" must be preserved. Models of a state-controlled and fragmented Internet are rejected.
2. The controversial issue of "enhanced cooperation", behind which some governments have been hiding efforts since 2005 to introduce international state control over the Internet, is circumvented with skilful wording and turned into something positive. Paragraph 108 states: "We reaffirm that Internet governance should continue to flow from the provisions set forth in the outcomes of the summits held in Geneva and Tunis, including in relation to enhanced cooperation."
3. The Internet Governance Forum (IGF) is described as a "unique platform for multistakeholder discussion on Internet governance issues" that has evolved from a one-off annual event into a complex ecosystem with over 170 national and regional IGFs. In contrast to the Elements Paper, the Zero Draft proposes not only to extend the mandate of the IGF, but also to give the IGF permanent status within the UN system (The IGF shall be made a permanent forum of the UN). It calls for further development of the "modalities" of the IGF, without however specifying this in detail. At the IGF in Lillestroem, there was discussion about whether a UNCSTD or MAG working group should be established to develop these specifications.
4. The Zero Draft remains unclear on the issue of IGF funding. Paragraph 118 calls for a strengthening of the IGF Secretariat in Geneva and asks UN Secretary-General Guterres to develop proposals. The Zero Draft does not say what these might look like or how the private sector could be involved.
The chapter on the future of the WSIS framework focuses on linking the eleven WSIS action lines more closely with the 16 SDGs and the implementation of the GDC recommendations. Synergies should be exploited and waste of resources through duplication prevented. It calls for the results of WSIS+20 to be given appropriate consideration at the SDG review conference scheduled for 2030.
The UN Group for the Information Society (UNGIS) will play a special role in implementing the WSIS action lines in the future. Around 40 UN organisations belong to UNGIS. UNGIS should develop into a "platform for multistakeholder dialogue" and be open to "multistakeholder advice". The proposal put forward during the consultations on the Elements Paper to institutionally supplement UNGIS by forming a "Stakeholder Advisory Group" (SAG) and to refer to the "Sao Paulo Multistakeholder Guidelines" (SPMGs) was not included in the Zero Draft. , the UNCSTD is also to be strengthened as a WSIS coordinating body and as a liaison to ECOSOC and the UN General Assembly (UNGA).
Another review conference (WSIS+30) is planned for 2035. For the high-level meeting planned for that date, further development of the multistakeholder approach is already being called for: "Involving the input and participation of all stakeholders, including in the preparatory process."

b. According to the WSIS+20 Preparatory Process Roadmap[3], comments on the zero draft were to be submitted to UNDESA by 3 October 2025. The roadmap also sets out the next steps on the way to the high-level meeting, which is planned for 16 and 17 December 2025 as part of the 80th UN General Assembly in New York.

A virtual consultation round with stakeholders on the zero draft is scheduled for 13 and 14 October 2025. Formal government negotiations will then begin in New York on 15 October 2025. It is still unclear how open these will be to stakeholders and what role the newly created Informal Multistakeholder Sounding Board (IMSB) will play in this process. On the sidelines of the ICANN meeting in Dublin, consultations with the co-facilitators from Kenya and Albania will take place on 27 and 28 October 2025.
In light of these consultations and the first round of negotiations, a revised draft of the final document is expected in early November 2025. This draft will be the subject of further stakeholder consultations and will form the basis for the final government negotiations (December 2025). According to the procedures for WSIS+20 adopted by the UNGA, it is incumbent upon the President of the UNGA, former German Foreign Minister Annalena Baerbock, to "collect" the "input" from the stakeholder consultations and ensure that this "input" is incorporated into the intergovernmental negotiations. How this will happen in concrete terms is still unclear.
The final document is to be adopted in New York on 17 December 2025 in the form of a UN resolution. The preceding two-day high-level conference is to take place at ministerial level. WSIS documents have so far been adopted by consensus. However, at the UNCSTD meeting in Geneva in April 2025, a WSIS resolution was voted on for the first time. The USA was the only country to vote "no". It rejected references in the WSIS resolution to the SDGs, gender and climate change. Votes on UN resolutions with "no" votes and abstentions are not uncommon in UNGA practice. Unlike the UN Security Council, no country has a veto right in the UNGA.

c. Governments have so far refrained from commenting on the zero draft. The final document of the Shanghai Cooperation Organisation (SCO) summit makes no mention of WSIS+20. However, the document contains an article calling for state regulation of „national Internet segments“[4]. Some governments , such as Switzerland, Australia and the United Kingdom, have responded positively to the zero draft. The EU has been the most vocal. On 10 September 2025, the European Parliament passed a resolution calling for permanent status for the IGF within the UN and backing the multistakeholder model for internet governance (to permanently renew the mandate of the IGF, and to strengthen its resources and the multistakeholder model of internet governance)[5]. EU Commissioner Hella Virkkunen has also advocated for a permanent IGF. [6] The negotiations beginning on 15 October 2025 in New York will be a first test of whether WSIS+20 will proceed largely without problems or become a platform for geopolitical controversies.

II. Artificial intelligence

The international debate on artificial intelligence in the third quarter of 2025 was dominated by two events in particular. On the one hand, at the end of August 2025, the UN finally agreed on the modalities for creating two new AI bodies within the UN – the AI Panel and the AI Governance Dialogue. On the other hand, the US and China presented their global AI strategies at the end of July 2025. The two strategies differ significantly and refer differently to developments within the UN. In addition, more and more countries and groups of countries (BRICS & G7) are announcing their AI strategies.

a. On 29 August 2025, the 79th UN General Assembly finally gave the green light for the creation of the two new UN bodies on AI with Resolution 79/302[7]: the Independent Scientific Panel on Artificial Intelligence/ISPAI (AI Panel) and the Global Dialogue on Artificial Intelligence Governance/GDAIG (AI Dialogue).

The AI Panel will consist of 40 members elected by the UN General Assembly on the recommendation of the UN Secretary-General for a term of three years.
1. Panel members will be elected "in their personal capacity" and not as representatives of a country or institution. However, there is a limit of no more than two representatives per country and company. Members of the AI Panel must submit a conflict of interest (COI) declaration to ensure the independence of the panel and prevent it from being captured by commercial or political groups. UN officials are excluded from membership. The panel will be led by two co-chairs, one from the North and one from the South. Working groups and advisory bodies may be formed.
2. The AI Panel is modelled on the UN Intergovernmental Panel on Climate Change. It will publish an annual comprehensive report on the state of AI development in the world, as well as thematic papers. The report will be based on "independence, scientific credibility, rigour, multidisciplinary and inclusive participation" and will provide comprehensive information on emerging opportunities and risks ( ). The report will serve as the basis for the AI Dialogue.
The AI Dialogue is to take place once a year for two days in conjunction with an existing UN conference.
1. The AI dialogue is to be organised in a multi-stakeholder format with a strong intergovernmental segment "as a platform to discuss international cooperation, share best practices and lessons learned, and to facilitate open, transparent and inclusive discussions on AI governance with a view to enabling AI to contribute to the implementation of the SDGs and to closing the digital divides between and within countries." The "social, economic, ethical, cultural, linguistic and technical implications" of AI are to be discussed. The dialogue should help to close capacity gaps, particularly between North and South, and work towards the interoperability of different AI governance approaches. Discussions will focus on robust oversight bodies for AI systems, their transparency and responsibilities, the use of open source, open data and open AI models, and respect for human rights.
2. On 25 September 2025, an informal kick-off of the AI Dialogue took place in New York in the presence of UN Secretary-General Guterres and UNGA President Baerbock.[8] The first AI Dialogue is planned for July 2026 in Geneva on the sidelines of the ITU's "AI for Good Summit". The President of the UNGA has yet to appoint two co-chairs who will determine the priorities through intergovernmental consultations. States are called upon to enable the participation of experts from the Global South through voluntary contributions. The second AI Dialogue is scheduled to take place on the sidelines of the UN Forum on Science, Technology and Innovation (STI Forum) in New York in June 2027. The continuation of the AI Dialogue will then be decided at the GDC Review Conference in 2027.
3. The proposal to link the AI Dialogue with the IGF was not accepted. At the kick-off event for the AI Dialogue on 25 September 2025, neither Guterres nor Baerbock made any reference to the IGF. This was again criticised in numerous comments on the WSIS+20 Zero Draft. There were calls for closer coordination between the IGF and its "IGF Policy Network on Artificial Intelligence“ [9] and the new UN AI Dialogue.

b. On 23 July 2025, US President Donald Trump presented his new global AI strategy entitled „Winning the Race: America´s AI Action Plan“[10] at the White House. Three days later, on 26 July 2025, Chinese Premier Li Qiang presented China's concept for a global strategy entitled „Action Plan on Global AI Governance“[11] at the World AI Conference (WAIC) in Shanghai.

The title of the US document, "Winning the Race," already makes the nature of the US strategy clear. The US wants to dominate the AI world of tomorrow. In the foreword, President Trump writes: "Breakthroughs in AI have the potential to reshape the global balance of power, spark entirely new industries, and revolutionise the way we live and work. As our global competitors race to exploit these technologies, it is a national security imperative for the US to achieve and maintain unquestioned and unchallenged global technological dominance." The paper contains three chapters: 1. AI Innovation, 2. AI Infrastructure, 3. AI Diplomacy:
1. The section on AI innovation deals with digital skills and the removal of regulatory and bureaucratic barriers to AI development. The US education system must be fundamentally restructured to meet AI requirements in order to have enough AI workers to secure the US's leading role. Regulations must not limit innovation. First, AI opportunities must be exploited, and only then should the risks of AI be discussed.
2. The section on AI infrastructure deals with the construction of AI factories and data centres, as well as the development and production of the latest generation of chips. It is necessary to establish a secure cyber infrastructure. "Secure by Design AI Technologies and Applications" must be developed.
3. The section on AI diplomacy aims to make American AI systems the global standard. "The US must also drive adoption of American AI systems, computing hardware, and standards throughout the world." The strategy offers like-minded countries close cooperation under US leadership. The document refers to initiatives by the UN, OECD, G7, G20, ITU and ICANN, but views them with scepticism. "Too many of these efforts have advocated for burdensome regulations, vague 'codes of conduct' that promote cultural agendas that do not align with US values, or have been influenced by Chinese companies attempting to shape standards for facial recognition and surveillance."
China's "Global AI Governance Action Plan" of 26 July 2025 takes a different approach. In its 13 chapters, it avoids a "China First" approach and calls for an expansion of equal international AI cooperation based on "global solidarity". The opportunities and risks of the technological revolution must be taken into account equally. It calls for an "inclusive, open, sustainable, fair, safe, and secure digital and intelligent future for all." AI must be developed "in service of humanity, respecting national sovereignty, aligning with development goals, ensuring safety and controllability, upholding fairness and inclusiveness, and fostering open cooperation."
1. Similar to the US paper, China is focusing on innovation, infrastructure and capacity development. It calls for an "innovation-friendly policy environment," strengthening "policy and regulatory coordination" and "the removal of technology barriers." Unlike the US paper, however, the Chinese plan takes the interests of developing countries into account, emphasising the need for equal cooperation and assistance in building diverse, open and innovative AI ecosystems that are also based on open source. Charles Mok described this in „Tech Policy Press“[12] as a "Chinese AI charm offensive." It does not mean that China is not striving for technological superiority and protecting its national security. However, this is not the primary concern of the action plan. The aim is to present the Chinese model as an alternative to "winning the race."
2. China places particular emphasis on the public sector. The protection of personal data is emphasised. However, freedom of expression and mass surveillance are not mentioned. The consequences of AI for the environment are taken into account. The aim is to develop low-power chips and efficient algorithms. Social ethics need to be strengthened. Standards should help to eliminate algorithm-based bias and discrimination. This is the task of standardisation organisations such as ITU, ISO and IEC.
3. As far as the role of the UN is concerned, the action plan supports the new UN initiatives for an AI panel and an AI dialogue. The goals set out in the GDC must be implemented, including through intensified cooperation between all stakeholders. Chapter 13 proposes the establishment of an inclusive multi-stakeholder AI governance model: "We support the establishment of inclusive governance platforms based on public interests and the joint participation of relevant entities. We encourage AI enterprises from different countries to engage in dialogue and exchanges, learn from each other's application practices in various fields of AI, and promote innovation, application, as well as ethical and safety cooperation in specific domains and scenarios."
4. The proposal for a new international multi-stakeholder AI organisation based in Shanghai caused quite a stir. What would such an organisation be responsible for and how would it relate to the two new UN institutions? Observers compared the plan to the dispute over DNS in the 1990s between the ITU and the US, which led to the establishment of ICANN, a private company based in Los Angeles. Leading private Chinese AI companies have already expressed interest in participating in such a new organisation.

c. On 6 July 2025, the BRICS countries adopted a "BRICS Leaders' Statement on the Global Governance of AI" at their summit in Rio de Janeiro. [13] In addition to the five founding members Brazil, Russia, India, China and South Africa, BRICS now includes six other countries (Saudi Arabia, Egypt, United Arab Emirates/UAE, Ethiopia, Indonesia and Iran) as well as ten observers (Belarus, Bolivia, Kazakhstan, Cuba, Malaysia, Thailand, Uganda, Uzbekistan, Nigeria and Vietnam).

The declaration contains a total of 21 recommendations in five chapters. The focus is on recommendations for strengthening national AI capacities, expanding infrastructure, and promoting innovation and further education. The younger generation must be offered sustainable and dignified employment opportunities. Developing countries should not be forced into the role of "data suppliers," but should cooperate on an equal footing with other countries for mutual benefit. "Digital sovereignty and the right to development are central to global AI governance," the statement says. And "collaborative governance of AI is complex, but possible."
The BRICS declaration supports a rules-based approach to AI. "Fair competition and market regulation" are central elements of an AI future. Barriers to access to AI technologies or financing must be removed, and the protection of intellectual property and public interests in developing countries must be strengthened. This must be reflected in internationally agreed technical standards. "We must avoid standard setting processes being used as barriers to market entry for smaller companies and developing countries."
The BRICS Declaration supports the GDC and the two new UN AI bodies: "The UN is central to global AI governance." Only the UN offers developing countries the opportunity to participate equally in international AI discussions. It warns against fragmentation of the AI governance discussion in regional or private bodies (OECD, G7, Bletchley Process). The UN is the only "fully inclusive and representative international framework."

d. The G7 countries (USA, Canada, Japan, Italy, France, Great Britain and Germany) already adopted an AI strategy at the G7 summit in Hiroshima in July 2023 (Hiroshima AI Process/HAIP)[14].

This strategy was further developed at the G7 summit in Kananaskis, Canada, in June 2025. The „G7 Leaderrs Statement on AI for Prosperity“[15] adopted on 17 June 2025 complements the „G7 Action Plan for a human centered adotion of safe, secure and trustworthy AI“[16] from 2024. This gives the G7 countries a comprehensive framework for AI development. They continue to rely primarily on the OECD as a platform and on strengthening the role of the private sector, but are also participating in the two new UN bodies on AI.
The G7's AI strategy is based on a human-centred approach. AI innovations must benefit people, minimise risks and strengthen national security. The new statement addresses topics such as the exploding energy demand for AI factories and data centres, as well as strengthening the public sector in the application of AI in areas such as education, health, transport and the environment. The consequences for culture and the creative industries were also discussed for the first time.
The partnership with emerging markets and developing country partners is to be strengthened. Investments in small and medium-sized enterprises and locally led AI-enabled innovation are to be promoted. Cooperation with universities in developing countries is to be strengthened. The document briefly refers to the GDC in its preamble, but avoids any reference to the UN's AI initiatives in its operative parts. No new financial resources were provided by the G7 countries.
A "G7 AI Adoption Roadmap" was agreed as an annex to the statement, which includes measures to strengthen small and medium-sized enterprises, measures to strengthen trust in AI applications and an exchange of talent. The US endorsed the statement but made it clear that its national strategy, which was adopted by the White House four weeks later, takes precedence over the HAIP.

III. Cybersecurity

Cybersecurity was on the agenda at several conferences in the third quarter of 2025. The final meeting of the Open-Ended Working Group (OEWG) took place in New York in July 2025 [17], , and the autumn meeting of the GGE LAWS on autonomous weapon systems was held in Geneva in September 2025 [18]. At the same time, preparations were made for the signing of the new UN Convention against Cybercrime. The signing ceremony is scheduled for 24 October 2025 in Hanoi. At the 60th session of the UN Human Rights Council in September 2025 in Geneva, a report entitled "Human rights implications of new and emerging technologies in the military domain" was discussed [19].

a. After five years, the OEWG concluded its work with the adoption of a final report on 27 July 2025 in New York.[20] Despite geostrategic tensions, the report was adopted by consensus, although some governments, including Russia, issued unilateral statements expressing dissenting views. The final report recommends transforming the OEWG into a so-called permanent "Global Mechanism" to continue discussing the five substantive issues that the OEWG had been addressing since 2020. Should the 1st UNGA Committee agree, the issue of cybersecurity would now have an institutional "home" within the UN system.

Overall, the OEWG's track record is mixed. The "Programme of Action" (PoA) demanded by Western states since 2021, which was intended to evaluate the implementation of the eleven basic principles for responsible state behaviour in cyberspace adopted in 2015, did not come to fruition. However, Russia's proposal to supplement the eleven norms with further norms and to convert them into a UN convention binding under international law was also rejected. Western states are not refusing to discuss further standards, but argue that clarity should first be established on how the existing standards, which are now widely recognised as the „UN Cybersecurity Framework“ [21], are being implemented before talking about new standards and making them binding under international law. The discussion is likely to continue to go round in circles in the new "Global Mechanism" for the time being. The OEWG final report does not contain a mandate for the drafting of a binding UN convention on cybersecurity.
The discussion on the application of international law in cyberspace also remains unsatisfactory. In 2013, a Group of Governmental Experts (GGE) reached agreement that international law, and in particular the seven jus cogens principles of the UN Charter, also apply to cyberspace, i.e. are relevant both offline and online.[22] Here, too, a PoA should determine how UN member states are implementing this agreement. In recent years, some governments have presented national position papers on how they practise the application of international law in cyberspace. However, there is no complete overview. The new "Global Mechanism" must also do some follow-up work here.
The discussion on confidence-building measures in cyberspace (CBMCs) produced a concrete result. After lengthy discussions, the 193 UN states agreed to create an international „Point of Contact Directory“ (PoC)[23]. The PoC is intended to function like the "red telephone" between the nuclear powers. Each UN member state will designate a contact person who can be contacted in the event of an incident affecting cyber security. The aim is to avoid misunderstandings, clarify unintentional cyber security breaches and de-escalate tensions. Such a PoC directory already exists within the OSCE for 53 states and has proven its worth there. The extension of this mechanism to 193 states has been widely welcomed. It remains to be seen how it will work in practice. In the first few weeks of the new PoC, it was already apparent that some states seemed to be testing the resilience of the system with a veritable wave of sometimes unfounded enquiries. An initial evaluation of the PoC mechanism is planned for 2026.
There was little progress on capacity-building measures. All 193 OEWG members agreed that there is a particular need for action in developing countries and that measures are needed to enable governments to protect their own cyber security. The debate centred on who should implement such measures, how the teaching programmes should be designed and who should finance them. In 2023, a round table was held for the first time, at which non-governmental stakeholders were represented on an equal footing with governments. According to the recommendations of the OEWG final report, the round table is to be established as a permanent body and supported by a Global ICT Security Cooperation and Capacity Building Portal (CCBP).
A constant point of contention within the OEWG was the participation of non-governmental experts from business, science, civil society and the technical community. Since 2021, OEWG Chair Burhan Gafoor from Singapore had been campaigning to make the voices of non-governmental stakeholders heard. Some governments rejected such participation. The compromise was that NGOs accredited by ECOSOC could participate in the OEWG deliberations. Non-governmental stakeholders who did not have ECOSOC accreditation could apply to participate, but required the approval of all 193 OEWG members. This gave every government a veto right against unwelcome NGOs. Russia in particular banned around 30 NGOs, including the World Economic Forum Davos (WEF) and the German Council on Foreign Relations (DGAP). Conversely, Ukraine vetoed the Moscow State Institute of International Relations (MGIMOI), among others. The new recommendations (Additional Elements on Modalities on the Participation of Other Interested Parties and Stakeholders, including Businesses, Non-Governmental Organisations and Academia) provide for a slight increase in flexibility. Governments still have the right of veto. However, if they exercise this right, they must justify it publicly and enter into consultations.
The OEWG final report will be discussed by the First Committee of the UNGA in October 2025. If adopted, the 80th UNGA would formally establish the Global Mechanism in a UN resolution in December 2025. The Global Mechanism consists of a plenary assembly with all 193 UN states and two thematic working groups. It is to meet twice a year and report to the UN General Assembly. The first meeting could take place in New York in March 2026.

b. The autumn session of the GGE LAWS took place in Geneva at the beginning of September 2025. The pressure on the GGE Laws to bring the negotiations, which have been dragging on for ten years, to a concrete conclusion had grown considerably in the run-up to the meeting. Against the backdrop of the further development of these weapons systems and their use in Ukraine, the Middle East and Africa, UN Secretary-General Guterres and the President of the International Red Cross are calling for an agreement to be concluded by the end of 2026.

On 6 September 2025, the Geneva meeting agreed on the essential elements of a "rolling text". The basic idea behind a possible agreement is a so-called "two-tier approach". On the one hand, autonomous weapon systems that are beyond human control and independently identify targets – objects or persons – in order to destroy or kill them are to be banned. On the other hand, the use of other autonomous weapon systems that use AI should be bound by the rules of international humanitarian law, as contained in the 1949 Geneva Conventions on the protection of civilians in armed conflicts.
Since 2023, thanks to an initiative from Austria, the issue has also been discussed in the UN General Assembly. In 2024, UN Secretary-General Guterres published a report with statements from more than 100 governments and NGOs.[24] The report was discussed in May 2025 during two days of informal consultations in New York.[25] In October 2025, the First Committee of the UNGA will consider the outcome and decide on the next steps.
Following the conclusion of the GGE LAWS negotiations in Geneva, Brazil published a statement on behalf of 41 states calling for the immediate commencement of formal negotiations on an international LAWS treaty. "The rolling text contains a set of elements that we consider to be a sufficient basis for negotiations on an instrument on lethal autonomous weapon systems."[26]

c. The UN Convention on Cybercrime [27], adopted by the 79th UN General Assembly in December 2024, will be officially opened for signature in Hanoi on 24 October 2025.[1] The Convention will enter into force 90 days after the deposit of the 40th instrument of ratification.

Critics of the convention complain about vague definitions of criminal offences, insufficient human rights safeguards in cross-border criminal prosecution, and the intention to draft a protocol dealing with the dissemination of criminal content in cyberspace. Western countries would have preferred all 193 UN member states to have acceded to the Council of Europe's Budapest Convention of 2001. This convention has clear definitions and protection mechanisms. However, many developing countries felt unrepresented by the Budapest Convention, as they were not involved in its drafting. To date, 71 countries have ratified the Budapest Convention.
The new UN Convention on Cybercrime does not contradict the Budapest Convention. States can accede to both conventions. Future practice will show which instrument is more effective. The EU- s have already expressed their intention to sign the new "Hanoi Convention". In December 2024, the US agreed to the convention, subject to ratification. It remains to be seen whether the new US administration will stand by this word. Sustainable support for the convention comes from the states of the Shanghai Cooperation Organisation (SCO) and BRICS.

d. At the 60th session of the UN Human Rights Council in Geneva in September 2025, a report entitled „Human rights implications of new and emerging technologies in the military domain“[29] was on the agenda. The 20-page report examines the possible consequences for human rights of AI applications in the military sphere. The report discusses scenarios relating to "lethal autonomous weapons systems, cognitive warfare and energy weapons". It examines the consequences of AI for the control of nuclear, biological, chemical and conventional weapons. It emphasises that governments have human rights obligations under international law and the Geneva Conventions that must be taken into account in the development and use of weapons systems. Seven recommendations remind governments of their responsibilities, while four recommendations are addressed to non-state stakeholders.

IV. State control over internet resources

In the third quarter of 2025, the discussion about possibilities for international state control over critical Internet resources flared up again. Russia put the issue on the agenda of the ITU Council Working Group on International Internet Related Public Policy Issues (CWG-Internet) in Geneva in September 2025. Indirectly, the issue was also debated at the SCO summit and various BRICS meetings, as well as at the WSIS+20 consultations.

a. For years, Russia has been trying to use the ITU, and in particular the CWG Internet, to raise the issue of state control over the management of critical Internet resources. Russian representatives point to alleged gaps in responsibility in existing mechanisms such as ICANN, RIRs and the IETF. In April 2025, Russian Deputy Foreign Minister Verschinin spoke of an "Internet management gap" in a conversation with UN Tech Envoy Amandeep Gil Singh [30]. At the Russian "Global Digital Dialogue" in Nizhny Novgorod in June 2025, the idea of creating a new intergovernmental body to oversee the management of domain names, IP addresses and Internet protocols was discussed once again.

b. At the 22nd meeting of the CWG Internet in September 2025, Russia proposed sending a questionnaire to all ITU member states to find out how widely accepted the established system of Internet resource management is. The aim was to examine whether the Internet should be defined as a "global public good", whether existing national and international legislation applies to the management of critical resources, and to what extent private organisations are able to act in the public interest. The question of whether international coordination is necessary at all was also raised: "Is it possible for States to manage the Internet at the national level without international coordination?“[31] The proposal did not receive majority support. Even potential supporters of Russia pointed to the "complexity" and limited mandate of the CWG-Internet. The issue could possibly be discussed at the next ITU General Assembly (ITU-PP 26) in Qatar in autumn 2026.[32]

c. However, Russia's proposal to establish a state supervisory mechanism for the management of critical Internet resources has met with little support, even among Russia's partners in BRICS and the SCO. Neither the SCO summit in Tijain in September 2025 nor the BRICS summit in Rio de Janeiro in June 2025 supported the proposal. However, the SCO declaration emphasises the principle of digital sovereignty. "It considers it important to ensure equal rights for all countries to regulate the internet and the sovereign right of states to manage it in their national segments."

d. At the final meeting of the OEWG at the end of July 2025, the Russian representative agreed to the final report but issued a unilateral statement reiterating Russia's demand for the development of legally binding norms for cyberspace. The new global mechanism must focus "on developing new legally binding norms in the field of digital security." At the same time, he criticised the participation of non-governmental stakeholders who, he said, were abusing their role for anti-Russian propaganda. "We insist that it is unacceptable to undermine the intergovernmental nature of negotiations on security in the field of ICT use within the framework of the Global Mechanism."[33]

e. During consultations on WSIS+20, Russian government representatives have so far refrained from calling for state supervision of Internet resources. Paragraph 104 of the Zero Draft explicitly advocates an open, free and global Internet and "rejects models of a state-controlled or fragmented Internet architecture." On 25 September 2025, Russian Deputy Foreign Minister Verschinin met with UN Tech Envoy Amandeep Gil Singh and ITU Secretary-General Doreen Bogdan Martin in New York. The communiqué on the meeting states that Russia will work "in a constructive manner to carry out the resolutions of the WSIS and to bridge the digital divide."[34]

V. Digital tax and customs duties

The economic dimension of internet governance returned to the agenda with greater force in the third quarter of 2025. For several years, discussions have been ongoing within the context of the G20/OECD and the World Trade Organisation (WTO) about a digital tax and customs duties for cross-border data trade.

a. After years of negotiations in the OECD/G20, the issue of digital taxation was resolved at the G20 summit in Rome in 2021 with the adoption of the OECD/G20 Inclusive Framework (IF). Pillar 2 of the agreement, also known as the BEPS (Base Erosion and Profit Shifting) Agreement, provided for a minimum tax rate of 15 per cent for tech companies. Under the first Trump administration, the US had briefly left the negotiations in 2020. The Biden administration later returned to the negotiating table and finally agreed to the BEPS Agreement.

In February 2025, under the new Trump administration, the US declared that it would not accept any regulations that disadvantaged American tech companies.[35] A memorandum entitled "Defending American Companies and Innovators From Overseas Extortion and Unfair Fines and Penalties" also threatened countermeasures.
The G7 countries then attempted to find a compromise that would keep the Inclusive Framework, which now includes 136 countries, alive while also making concessions to the US. The proposal of 28 June 2025 is a so-called "side-by-side" model with exceptions for US companies ("side-by-side" solution under which US-parented groups would be exempt from the Income Inclusion Rule (IIR) and Undertaxed Profits Rule (UTPR) in recognition of the existing US minimum tax rules to which they are subject).. The G7 hopes that this compromise will contribute to a constructive dialogue on taxation in the digital economy and the safeguarding of the tax sovereignty of sovereign states.[36]
The reactions of the other G20 members varied. In August 2025, developing countries in particular initiated a new round of negotiations on global digital taxation within the framework of the UN.[37] The aim is to establish a UN Framework Convention on International Tax Cooperation. The newly formed International Negotiations Committee (INC) met for its first two sessions in New York in August 2025 and identified issues and priorities. A third round of negotiations is scheduled for November 2025 in Nairobi. Three further rounds are to follow in New York and Nairobi in 2026. The convention is expected to be finalised by the end of 2027.
At the meeting of G20 finance ministers in South Africa (KwaZulu-Natal) on 18 July 2025, attempts were made to play down the conflict. The final declaration continues to support the implementation of the Inclusive Framework (IF) with the BEPS agreements, but takes note of the negotiations that are beginning within the framework of the UN. The exemptions for US tech companies proposed by the G7 are ignored in the statement and are not mentioned. “We also welcome discussions to enhance the effectiveness and inclusivity of the IF. We note the ongoing negotiations to establish a United Nations Framework Convention on International Tax Cooperation.“[38]

b. The issue of tariffs on cross-border trade in digital services has been pending since the late 1990s. Based on a proposal by the then US administration under Bill Clinton, a moratorium on imposing tariffs was agreed. In the 2010s, more and more developing countries called for an end to this moratorium, hoping to generate additional revenue.

In 2019, 79 countries launched a joint initiative on the sidelines of the World Economic Forum (WEF) in Davos to start negotiations on trade-related aspects of electronic commerce with the aim of establishing a permanent agreement (Digital Data Trade Pact).[39] The aim should be to preserve the basic idea of the moratorium, which is to enable the free flow of data services, while taking into account the specific interests of developing countries. Over 90 governments participated in the negotiations within the framework of the WTO.
In July 2024, negotiators from Singapore, Japan and Australia announced that key parts of the agreement had been negotiated.[40] „After five years of negotiations under the WTO Joint Statement Initiative on Electronic Commerce, participants have reached a new phase, achieving stabilised text on the attached Agreement on Electronic Commerce that reflects a balanced and inclusive outcome.“ The plan was to sign the agreement before the next WTO Ministerial Conference/MC14 (March 2026 in Yaounde, Cameroon). Negotiations have now stalled again. The moratorium, which is to be replaced by the new treaty, ends on 31 December 2026. At a WTO meeting on 25 September 2025 in Geneva[41], Jamaican Ambassador Richard Brown, facilitator of the "WTO Work Programme on Electronic Commerce", assured that a tangible outcome "for the stability and development of digital trade" remains a priority for MC14.