Developments in the internet governance environment July to August 2026

Wolfgang Kleinwächter Sep 8, 2026 12 min read

UN AI Panel Releases First Report

On 1 July 2026, the UN’s ‘Independent International Scientific Panel on AI’ (IISPAI) published its first report. Spanning 50 pages, it addresses AI developments across seven areas: AI research; health, education, science and agriculture; economic impacts; security and the environment; human rights and democracy; culture, personal freedom and child protection; and management and governance. The 40-member panel, chaired by Turing Award winner Yoshua Bengio and Nobel Peace Prize laureate Maria Ressa, highlights a dilemma: “Policymakers need evidence to make informed, consequential governance decisions, but by the time the evidence exists, it might be too late to make them, as the evidence lags behind the pace of AI development.” The new UN AI Council is modelled on the UN Intergovernmental Panel on Climate Change and is intended to provide governments with facts so that they can make appropriate decisions. The UN report was presented at the first ‘Global Dialogue on AI Governance’ on 6 and 7 July 2026 in Geneva. A second report is due to follow in April 2027.[1]

Geneva Seeks Global AI Consensus

The first ‘Global Dialogue on AI Governance’ took place in Geneva on 6 and 7 July 2026 to discuss the report by the UN AI Panel.[2] The focus was on seeking an international consensus regarding AI testing, AI risks and AI standards, particularly in relation to child protection. A key demand was that humans must retain the final decision-making authority in critical areas such as law enforcement, healthcare, the labour market and military AI applications. UN Secretary-General Guterres spoke out in favour of global AI regulations: “Some might claim that governance is the enemy of innovation. But innovation needs guardrails. The technologies we trust most—in aviation, in medicine, in nuclear energy and beyond—earned that trust because we acted to hold their makers to account. If AI is to be powerful, it must be governed. If AI is to be trusted, those who build it must be accountable. If AI is to be global, it must be fair. And if AI is to serve the future, it must not consume the future.”[3] Other topics included preventing the concentration and monopolisation of AI development and the role of the Global South. The fragmentation of discussion platforms was viewed critically, and a “dialogue of dialogues” was called for. The US government rejects the UN dialogue on AI governance. The second dialogue on AI governance will take place in early May 2027 on the sidelines of the UN STI Forum in New York.

ITU Launches AI for Good Global Commission

On 2 July 2026, the ITU announced the establishment of a new ‘AI for Good Global Commission’.[4] It is chaired by Rwanda’s President Paul Kagame and multi-billionaire Marc Benioff, CEO of Salesforce. The new “Global Commission” follows on from the “ITU/UNESCO Broadband Commission for Sustainable Development”, which has published an annual report on the global development of digital connectivity since 2010. The new commission includes, amongst others, the Presidents of Estonia and Iceland, Alar Karis and Halla Tómasdóttir; the Directors-General of the WTO, WIPO, UNESCO and UNDP; and numerous representatives from the private sector such as Andy Jassy (Amazon), Jack Clark (Anthropic), Albert Bourla (Pfizer), Brad Smith (Microsoft), Jensen Huang (NVIDIA), Margherita Della Valle (Vodafone), Xu Ziyang (ZTE), James Manyika (Alphabet) and Cristiano Amon (Qualcomm). “A key focus of the AI for Good Global Commission will be to bridge digital divides and help ensure that AI becomes a tool for solving global challenges, not deepening inequalities”. The Commission intends to advocate for a multi-stakeholder approach to AI governance, but representatives of civil society are not included. The relationship between the new ITU Commission and the new UN AI Panel also remains unclear.

China Founds Counterweight: WAICO in Shanghai

On 16 July 2026, a new intergovernmental organisation, the ‘World Artificial Intelligence Cooperation Organisation (WAICO)’, was founded in Shanghai.[5] The nine-page treaty was signed by 29 countries, including Brazil, South Africa, Kenya, Venezuela, Russia and Serbia. The WAICO treaty does not provide for a role for non-state representatives. WAICO aims to work towards ensuring “that AI is safe, fair, ethical, and beneficial to all humanity”. WAICO intends to be guided by the principles of the UN Charter and, in particular, to help countries in the Global South develop independent AI infrastructures. Western states have not joined WAICO . WAICO is headquartered in Shanghai[6]. By the end of August, the number of WAICO members had risen to 38. Commentators view the founding of WAICO as a Chinese response to the ‘Pax Silica’, established by the US in 2024, to which 25 states now belong.[7]

Xi Jinping Warns Against AI Monopolies

WAICO was founded on the eve of the “World AI Conference” (WAIC)[8]. At the opening ceremony on 17 July 2026, Chinese President Xi Jiping warned against AI monopolisation. He called for answers to four questions: “How to get along with thinking machines? How to ensure security when algorithms are part of decision-making? How to tackle ethical challenges posed by technology through adaptive governance? How can we make AI accessible to all when the digital divide continues to widen?” These must be guided by four principles: promoting innovation, assessing risk impacts, inclusion and solidarity. WAIC is the ideal instrument for this. “AI development should not be a solo performance by a single country but a symphony of international cooperation,” said Xi[9]. UN Secretary-General Antonio Guterres emphasised in his speech the central role of the UN in AI governance. “The Global Digital Compact (GDC) made it clear: technology that will shape the future of humanity must be shaped by all of humanity. It cannot be governed by a handful of countries or a handful of companies. Every nation needs a seat at the table. And the table has now been built through the Global Dialogue on AI Governance“[10]. UN Tech Envoy Amandeep Gil Singh was also in Shanghai.

WSIS Forum 2026: Bridging the Digital Divide

The WSIS Forum 2026 took place in Geneva from 6 to 10 July 2026[11]. It formed part of ‘Geneva Digital Week’ (which included the ITU’s AI for Good Summit and the Global Dialogue on AI Governance), bringing together a total of 12,000 participants from 177 countries. Discussions centred on the outcomes of the WSIS+20 Review Conference and the strategy up to 2035 (WSIS+30). The future of the IGF, which now has permanent status within the UN system, played a key role. The focus was on bridging the digital divide, artificial intelligence (AI) and the implementation of the WSIS Action Lines. 1,600 projects had been submitted for the WSIS Prize 2026. Of these, 270 qualified for the final round - including the ‘European Summer School on Internet Governance’ (EURO-SSIG) in Meissen - from which 17 winners were selected.

IGF-MAG Paves the Way to Nairobi

Under the new leadership of Jennifer Chung, the IGF-MAG held a total of four meetings in July and August 2026, paving the way for the 21st IGF in Nairobi in December 2026. At the same time, the groundwork was laid for strategic planning up to 2030. A new development is the idea of so-called Internet Governance Labs (IGLs), introduced by the EU. IGLs are a new format designed to examine how new technologies might affect the use of the Internet and its governance. The topic is controversial because the distinction between IGLs and the existing Policy Networks (PNs) and Dynamic Coalitions (DCs) is unclear. Proposals for IGLs can be submitted until 15 September 2026. On 25 August 2026, the IGF Secretariat published the 728 proposals for IGF 2026 workshops, Open Forum sessions, Day Zero events, lightning talks, networking sessions and awards submissions[12]. The IGF programme will be discussed at the 2nd Open MAG Consultations in Geneva in early September 2026. To ensure greater transparency, the MAG Chair has introduced so-called public consultation sessions open to the broader public (IGF Office Hours). The first took place at the end of August 2026.

New UN Cybersecurity Mechanism Takes Off

From 20 to 24 July 2026, the first meeting of the new, permanent UN Global Mechanism for Cybersecurity (G-MEC) took place in New York. G-MEC continues the five-year work of the OEWG. The new Chair is Egriselda López, El Salvador’s Ambassador to the UN, who is also a co-facilitator of the Global Dialogue on AI Governance. Discussions centred on the implementation of the eleven principles for responsible state behaviour in cyberspace, adopted in 2015; the implementation of the global directory of points of contact (POC), established in 2024; and the development of a system for digital capacity-building, particularly in the Global South. The involvement of non-state representatives remains a contentious issue. They are to play a greater role in the work of the two Dedicated Thematic Groups (DTGs), which are due to meet for the first time in December 2026 in New York to discuss ‘General Threats’ and ‘Capacity Building’. The long-standing dispute over whether the UN should focus on implementing the existing eleven non-binding cybersecurity standards or on drafting a new UN convention on cybersecurity that is legally binding under international law remained unresolved.

EU AI Act Fully Enters Into Force

On 31 July 2026, the EU’s AI Act came into force. The ‘Guidelines on transparency obligations for providers and deployers of certain AI systems’ had already been published on 20 July 2026. On 27 July 2026 , the ‘AI Omnibus’ came into force, simplifying some of the rules set out in the AI Regulation.[13] From 2 August 2026, the European Commission’s new AI Office will be responsible for ensuring compliance with the AI Act, particularly in relation to general-purpose AI models (GPAI). The AI Office is supported by national supervisory authorities (in Germany, the Federal Network Agency). Under the new rules, chatbots and other interactive AI systems must inform users that they are interacting with AI rather than a human. Deepfakes must be labelled. AI-generated or modified content must carry machine-readable labels. These measures are intended to reduce deception and manipulation. The highest governing body under the EU legislation is the ‘AI Board’, established in 2024, comprising all 27 EU Member States. It is advised by a panel of experts, which includes, from Germany, Mario Fritz (Helmholtz Centre Saarbrücken), Judith Simon (University of Hamburg) and Marius Hobbhan (Apollo Research).[14]

Trump Authorizes Private Sector Cyber Counterstrikes

On 12 August 2026, US President Trump issued a ‘Presidential Memorandum’ entitled ‘Expanding Capabilities to Combat Transnational Cyber-Enabled Crime’, granting US private companies greater powers to respond offensively to cyber attacks.[15] The memorandum authorises vetted US companies to carry out offensive cyber operations against foreign criminal groups. This applies in particular to responses to ransomware, phishing, financial fraud and sextortion attacks. Operations require prior written authorisation from the US government. Commentators have compared the decree to the system of letters of marque, under which, in the Middle Ages, pirates were authorised by kings to attack enemy ships.

AI Workers Warn of Losing Control

On 20 July 2026, 1,384 employees of leading AI companies in the US, including Anthropic, OpenAI, Microsoft, Meta, Alphabet, Accenture and Nvidia, articulated their concern in a statement entitled ‘Pacing the Frontier’ that the latest AI developments could spiral out of control. “There is a real risk that capability development will rapidly accelerate beyond our ability to understand or control the resulting systems.” They are calling on the US government to join international efforts to develop tools with which the risks posed by AI can be managed. “We request that the US government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development.”[16]

NATO and Accenture Build Secure Network

On 7 July 2026, on the sidelines of its ‘Summit Defence Industry Forum’, NATO signed a contract with the AI company Accenture to establish a Protected Business Network (PBN). This is intended to advance NATO on its path towards becoming a ‘digitally enabled Alliance’: the contract is worth US$200 million. ‘The PBN establishes the foundation for classified digital operations across the NATO Enterprise, ensuring that decision-makers and warfighters across domains can communicate, coordinate and access critical data in a modern, standardised and scalable cloud environment that is more resilient to external attack and disruption’.[17]

Germany and France Join Forces on AI Safety

On 17 July 2026, Federal Minister for Digital Affairs Wildberger and his French counterpart Anne Le Hénanff agreed, on the sidelines of the Franco-German Ministerial Council, on cooperation between the new AI security institutes: the “Institut national pour l’évaluation et la sécurité de l’intelligence artificielle (INESIA)” and the Federal Government’s AI Security Institute (AISI) in Berlin.[18]

ICANN: Record Number of gTLD Applications

On 12 August 2026, the application phase for ICANN’s new gTLD programme came to an end. On 13 August 2026, ICANN announced that it had received over 1,600 applications for new gTLDs.[19]. However, only 56 of these applications were supported through ICANN’s Applicant Support Programme (ASP)[20]. Through the ASP, ICANN aims to prevent a widening of the North-South divide in the global DNS market. The largest applicant is the Link Freedom Group, which has submitted applications for 316 new gTLDs.[21] The complete list of all applications will be published at on ‘Reveal Day’[22], possibly even before the next ICANN meeting in Bali (October 2026).

AFRINIC Names New CEO After Crisis

On 24 August 2027, the Board of Directors of AFRINIC, the Regional Internet Registry (RIR) for Africa, appointed former ICANN Director Mike Silver as its new CEO.[23] Silver was also Vice-Chair of the Address Supporting Organisation (ASO) at ICANN and is regarded as a renowned expert. After AFRINIC had been thrown into disarray by a series of management errors, there is now hope that Silver’s appointment will steer AFRINIC back into calm waters. AFRINIC had been drawn into a series of legal proceedings by the Hong Kong businessman Lu, CEO of Cloud Innovation, over the misuse of IPv4 address blocks; these proceedings led to the dissolution of the AFRINIC Board in 2024 and raised the question of the long-term viability of the self-governance model practised by ICANN for the management of IP addresses. Both the African Union (AU), with which AFRINIC is linked through a Memorandum of Understanding (MoU), and the Government of Mauritius, under whose jurisdiction AFRINIC operates, had intervened in the proceedings, and calls for state oversight of IP address management grew louder.[24] AFRINIC is integrated into the ICANN system as one of the five Regional Internet Registries (RIRs). Following the election of a new AFRINIC Board in 2026, the nomination of Mike Silver is a further stabilising factor, demonstrating that the multi-stakeholder model of Internet self-regulation works and that crises can be managed by the community itself without external interference. Mike Silver will take up his post on 1 January 2027.

Cybercrime in Africa Doubles

Africa is increasingly becoming a hotbed of cybercrime. As Interpol reports in its ‘African Cyberthreat Assessment Report 2026’, losses caused by cybercrime have more than doubled since 2024: from US$192 million to US$484 million. 55 per cent of these attacks are AI-based. There are now 1.2 billion mobile phones in Africa.[25]

ISOC Defends Multistakeholder Model Ahead of ITU Plenipotentiary

On 11 August 2026, ISOC published a position paper on the upcoming ITU Plenipotentiary Conference, scheduled for November 2026 in Doha. The agenda for ITU Plenipotentiary Conference 26 includes, amongst other things, proposals that could have implications for the multistakeholder model of internet governance. These include the idea of making a fresh attempt, following the failure of the WCIT (Dubai 2012), to extend the 1988 International Telecommunication Regulations (ITR) to the Internet. Furthermore, Russia has announced in the ITU Council Working Group on the Internet (CWG-Internet) that it intends to raise once again the issue of state oversight of the management of critical Internet resources. The ISOC paper states, amongst other things: “For the Internet Society, it is key to ensure that PP-26 upholds the existing multistakeholder model of Internet governance, where governments, the technical community, civil society, academia and the private sector all contribute according to their expertise, rather than expanding the ITU’s mandate into areas currently managed by multistakeholder organisations within the ecosystem. It is this model that has enabled the Internet to thrive.”

[1] https://www.un.org/independent-international-scientific-panel-ai/en/preliminary-report

[2] https://www.un.org/global-dialogue-ai-governance/en

[3] https://transcripts.un.org/en/asset/k1y/k1yd3dlebq

[4] https://www.itu.int/en/mediacentre/Pages/PR-2026-07-02-AI-for-Good-Global-Commission.aspx

[5] https://en.wikipedia.org/wiki/World_Artificial_Intelligence_Cooperation_Organization

[6] https://www.stdaily.com/web/English/2026-07/19/content_549692.html

[7] https://paxsilica.org/

[8] https://www.worldaic.com.cn/news/a7a32ea4f41f4e86937f51ec201e84b4?tab=0

[9] https://news.cgtn.com/news/2026-07-17/Full-text-Xi-s-keynote-speech-at-the-2026-WAIC-opening-ceremony-1OQSfeoRvUs/p.html

[10] https://unsdg.un.org/latest/announcements/technology-must-serve-people-not-other-way-around-un-secretary-general-world

[11] https://www.itu.int/net4/wsis/forum/2026/en/Home/Outcomes

[12] https://intgovforum.org/en/content/igf-2026-received-sessions