Developments in the internet governance environment in September 2026

Wolfgang Kleinwächter Oct 21, 2026 11 min read

SCO Summit in Bishkek: Call for equal rights in Internet regulation

On 1 September 2026, the 25th Summit of the Shanghai Cooperation Organisation (SCO) took place in Bishkek. Among those present were the Presidents of Russia (Putin), China (Xi) and India (Modi). In the final document[1] the heads of state and government warned against the militarisation of cyberspace and called for equal rights in the international regulation of the Internet, including for countries in the Global South. They called for the signing of the Hanoi Convention against Cybercrime. However, there was no mention of joining the World Artificial Intelligence Cooperation Organisation (WAICO), which was founded in Shanghai at the end of July 2026 under Chinese leadership.

GGE LAWS adopts consensus report on autonomous weapon systems

From 1 to 5 September 2026, the working group on autonomous weapon systems (GGE LAWS) held what is provisionally its final meeting in Geneva. It adopted by consensus a report containing ‘elements’ of a possible future treaty.[2] The report also contains a definition of autonomous weapon systems. It will now be submitted to the CCW Conference of States Parties, which is due to meet in Geneva in November 2026 and must decide on the formal commencement of negotiations for an agreement on autonomous weapons that is binding under international law.

G20 digital ministers agree on an AI education pact and the ‘Carolina Principles’

On 2 September 2026, the G20 digital ministers met in Chapel Hill, North Carolina.[3] The meeting developed into an ‘education summit’. The main focus was on how the world’s population can be prepared for the impending upheavals in the world of work brought about by AI. In a “G20 AI Prosperity Compact”[4], the ministers agreed on a wide range of educational measures. As the host, the US had pressed for agreement on principles for dealing with new technologies. The “G20 Carolina Principles for Emerging Technologies”[5] contain 24 recommendations designed specifically to promote innovation. A flexible approach to AI regulation is proposed.

ITU: World Telecommunication Policy Forum adopts five recommendations

The 7th ITU World Telecommunication Policy Forum (WTPF) took place in Nassau, the Bahamas, from 2 to 5 September 2026. It adopted five so-called ‘Opinions’ on the topics of ‘Strengthening ICT-centric innovation ecosystems and entrepreneurship’, ‘Space connectivity’, ‘Resilience of telecommunications/ICTs’, ‘Green digital transformation: climate change and environmental sustainability’ and ‘Bridging digital divides, particularly on gender and age as well as skills and connectivity’.[6] The recommendations will be submitted to the ITU General Assembly, which is due to take place in Doha in November 2026.

MAG consultations: IGF 2026 finalises workshops and main sessions

From 7 to 9 September 2026, the 2nd Open Consultations of the IGF’s Multistakeholder Advisory Group (MAG) took place in Geneva[7]. The main focus was on preparations for the IGF in Nairobi in December 2026. Out of the 456 proposals for workshops, 80 were selected, grouped into five thematic areas: 1. Digital sustainability, 2. Governance, 3. Access, 4. Human rights and 5. Cybersecurity. The four topics for the main sessions are AI governance, social media, digital infrastructure and digital inclusion. A working group (Kenya, Egypt, Australia, the United Kingdom, Pakistan, Switzerland) was formed to address the contentious issue of the role of governments in the IGF and the new ‘Governmental Dialogue’.

UNGIS Task Force discusses financing mechanisms for digital development

On 8 September 2026, the first meeting of the ‘UNGIS Task Force on Strengthening Financial Mechanisms for Digital Development’[8] took place in Geneva. The financing of digital development was one of the most contentious issues at the WSIS+20 Review in December 2025. The Task Force, chaired by the ITU, comprises 29 members, including representatives from the World Bank. It is required to put forward proposals for a financial mechanism for digital development in the Global South by 2027.

UNCSTD Working Group formulates initial principles for global data governance

On 8 and 9 September 2026, elements of a final report were discussed in Geneva at the seventh session of the UNCSTD Working Group on Data Governance (WGDG)[9]. The recommendations put forward so far are very general. Data governance is described as the foundation for AI governance. Nine very general principles have been formulated, which are essentially drawn from documents on internet governance. A two-tier data governance system is proposed, comprising universal principles at the global level and specific rules at the national level. The working group was established by the Global Digital Compact (GDC) and is required to report to the 81st UN General Assembly by the end of the year. The final meeting is scheduled for November 2026 in Geneva.

Anthropic CEO Amodei calls for global AI guidelines in a sensational essay

On 10 September 2026, Dario Amodei, CEO of Anthropic, published an essay entitled ‘We must Pace the Frontier’[1o]. In the essay, he warns against unchecked AI development. AI models could take on a life of their own, fall into the hands of terrorists and cause harm to humanity. He argued that “guardrails” are necessary. He acknowledged that AI developers are moving faster than regulators, particularly at an international level. He proposed a series of mechanisms, such as verification procedures during testing and external evaluators. Amodei received support for the essay from his rival Sam Altmann, CEO of OpenAI, and from Elon Musk.

20th BRICS Summit emphasises the UN’s role and the expansion of cyber security

On 12 September 2026, the 20th BRICS Summit took place in New Delhi. BRICS now has 10 members. The new members are Egypt, Ethiopia, Iran, Indonesia and the United Arab Emirates (UAE). All delegations were represented by their presidents. However, Brazil’s Lula sent his foreign minister instead due to the election campaign. The 45-page final declaration[11] includes, amongst other things, statements on cyber security, the digital economy and AI. The central role of the UN is emphasised across all these areas. A call is made for the signing of the Hanoi Convention against Cybercrime. Support is given to the drafting of an additional protocol to the Convention, which is intended to address criminal content on the Internet. On the subject of AI, the “BRICS Leaders’ Statement on the Global Governance of AI” from the 2025 BRICS Summit in Rio de Janeiro is reaffirmed.

China’s Minister of State Security warns against cognitive warfare via AI

On 14 September 2026, Chen Yixin, China’s Minister of State Security, published an article in the specialist journal *China Cyberspace* in which he warned that AI applications could undermine China’s national security: “Various hostile forces and malicious individuals could misuse generative AI technologies – such as deepfake audio and video recordings, AI-generated images and text, as well as intelligent online armies – to produce political rumours on a massive scale.” These tools could “spread harmful information, stir up hostile sentiments, and wage a war for the broader public and cognitive warfare.” Yixin cited AI-driven attacks on infrastructure, data theft and fundamental changes in the conduct of warfare as further threats. [12]

State of the Union address: Von der Leyen urges caution in AI development

On 16 September 2026, the President of the European Commission, Ursula von der Leyen, delivered her annual ‘State of the Union’ address in Strasbourg. She devoted much of her speech to the development of AI. She said she took the warnings from US AI companies about the risks of unchecked AI development seriously and proposed ‘shifting down a gear’. In November, von der Leyen plans to invite leading AI labs to discuss a joint strategy on model evaluation, verification, early warning and AI safety. By November 2026, the European Commission aims to present “pioneering initiatives” for the sectors of health, transport, agriculture and food, advanced manufacturing, as well as defence and space. Von der Leyen said: “We want AI. We want AI to create added value in a responsible manner. And to do so at lower cost and with lower energy consumption. But above all, we want AI where humans remain in control. That is the essence of the European approach.”[13]

US politics: President Trump announces the creation of a military ‘AI Force’

On 21 September 2026, US President Trump announced in Washington the establishment of an ‘AI Force’. This new unit is to be modelled on the military ‘Space Force’ and will coordinate the various AI activities of ministries and other government bodies.[14]

GDC Review kicks off on ‘Digital Development Day’ in New York

On 21 September 2026, the review of the Global Digital Compact began in New York with ‘Digital Development Day’. UN Deputy Secretary-General Amandeep Gil Singh, Director of the UN Office for Emerging Technologies (ODET), announced a comprehensive analysis of how the five objectives of the GDC are being implemented. includes, amongst others, the Working Group on Data Governance, the AI Panel and the Global Dialogue on AI Governance. The AI Panel presented its first “Thematic Brief”[15] entitled “AI Agents, Misalignment and the Risk of Losing Human Control: Evidence from the OpenAI-Hugging Face Incident”. The GDC review must be completed by the end of 2027.[16]

81st UN General Assembly: Growing geopolitical divide over AI regulation

On 22 September 2026, the 81st UN General Assembly began in New York. During the general debate, 128 states spoke on the subject of AI. At the opening, UN Secretary-General Guterres reiterated his warnings about unregulated AI development and proposed an ‘AI Risk Management Framework’[17]. US President Trump rejected any form of AI regulation. He proposed renaming “Artificial Intelligence (AI)” as “Super Intelligence” (SI)[18]. The heads of government of France (Macron[19]) and the UK (Burham[20]) spoke in favour of AI regulation. China’s Vice-President Han Zheng called for a “human-centred framework” to bridge the digital divide. Speakers from the Global South expressed fears of “digital re-colonisation”.

22 heads of state sign a call for the regulation of frontier AI models

On 22 September 2026, 22 heads of state and government published a “Call for Control of Frontier AI Models”.[21] The call was initiated by Norway’s Prime Minister (Store) and the President of Finland (Stubb). The call states, amongst other things: “AI must remain under human direction, oversight and control. It must be developed and used in line with international law.” It proposes considering the establishment of an international AI organisation capable of agreeing on binding universal standards and performing a supervisory role. The model is the International Atomic Energy Agency (IAEA) in Vienna. The appeal was signed by, amongst others, German Chancellor Merz and European Commission President von der Leyen. The heads of state of Canada and Australia, as well as those of Kenya, Kazakhstan, the United Arab Emirates, South Africa and Singapore, are also among the signatories.

Historic special session: UN Security Council debates artificial intelligence for the first time

On 23 September 2026, the UN Security Council in New York held a special session on AI for the first time in its history.[22] The experts invited were Joshua Bengio, co-chair of the UN’s AI panel and professor at the University of Montreal, as well as the CEOs of OpenAI, Sam Altman; Anthropic, Damio Amodei; and Hugging Face, Clément Delangue. Hugging Face fell victim to an autonomous hacker attack carried out by an AI developed by OpenAI. The experts warned of the risks posed by AI and called for regulations governing the development and use of AI. The majority of UN Security Council members also called for effective control of AI risks. The US rejected any form of AI governance[23]. Russia stated that the US Security Council had no jurisdiction over AI.[24]

US-China summit: Agreement reached on a ‘red telephone’ for serious AI incidents

On 23 September 2026, at the Sino-American summit[25] between Presidents Xi and Trump in Washington, it was agreed to establish a dedicated communication channel for serious AI incidents affecting national security. The model for this is the ‘red telephone’ between the Soviet Union and the US, which was used to prevent a nuclear war in the 1960s. A ‘US-China Superintelligence Dialogue’ was also agreed, which is due to meet for the first time in November 2026. [26]

‘White House Accord on Super-Intelligence’ relies on corporate self-regulation of AI

On 29 September 2026, President Trump and six of the leading AI companies (OpenAI, Anthropic, Meta, Google, Nvidia and Elon Musk) agreed in Washington on a “White House Accord on Super-Intelligence”.[27] The agreement focuses on self-regulation of AI. The companies undertake to introduce their own internal review procedures to keep AI risks under control. An “independent external auditor” is to be appointed. Alex Carp, CEO of Palantir, also attended the meeting.[28] The agreement was described by Democratic Senator Richard Blumenthal as futile and dangerous. The debate surrounding social media had shown that self-regulation does not work: “No public accountability, completely discretionary and a false sense of security”.[29]

Trump’s executive order aims to replace the term ‘Artificial Intelligence’ with ‘Super Intelligence’

On 29 September 2026, President Trump signed an executive order entitled ‘Inaugurating the Era of Super Intelligence (SI)’[30]. The executive order stipulates that the term ‘AI’ must be replaced by ‘SI’ in future. Michael Kraitsos, Director of the Office of Technology Development at the White House, is tasked with presenting a legally compliant definition of SI within 60 days.

California’s Governor Newsom defies Trump and sticks to the term ‘Artificial Intelligence’

On 30 September 2026, the Governor of California, Gavin Newsom, issued an executive order stipulating that the use of the term ‘Artificial Intelligence’ is legally binding in California[31]. Newsom had described Trump’s executive order on SI as ‘bullshit’. Most AI companies in the US are based in California.

OSCE conference in Switzerland: cyber diplomacy for de-escalation and peacekeeping

On 29 and 30 September 2026, an OSCE conference entitled ‘De-escalation in Cyberspace: Mediation and Preventive Diplomacy’ took place in Zug. As the OSCE Chair-in-Office, Switzerland sought to discuss, ahead of the OSCE Ministerial Meeting, ways in which cyber diplomacy can contribute to peace and security. Swiss Foreign Minister Ignaz Cassio summarised the outcomes as follows: “Technology can help us identify threats. Artificial intelligence can help us analyse them. But neither can decide when communication is appropriate rather than retaliation, or when de-escalation is needed rather than escalation. That remains our responsibility.“[32]

[1] https://eng.sectsco.org/images/07ea/09/0c/2512843.pdf

[2] https://docs-library.unoda.org/Convention_on_Certain_Conventional_Weapons_-Group_of_Governmental_Experts_on_Lethal_Autonomous_Weapons_Systems_(2026)/CCW-GGE.1-2026-3-Advance_copy.pdf

[3] https://www.g20.utoronto.ca/2026/260902-innovation-statement.html

[4] https://www.g20.utoronto.ca/2026/260902-ai-prosperity-objectives.html

[5] https://www.g20.utoronto.ca/2026/260902-carolina-principles.html

[6] https://www.itu.int/hub/2026/08/wtpf-26-which-key-questions-will-shape-the-digital-future/

[7] https://intgovforum.org/en/filedepot_download/342/31949

[8] https://www.itu.int/net4/wsis/ungis/Home/TaskForce

[9] https://unctad.org/meeting/7th-meeting-un-cstd-multi-stakeholder-working-group-data-governance-all-levels

[10] https://darioamodei.com/post/we-must-pace-the-frontier

[11] https://www.mea.gov.in/bilateral-documents?dtl/41776/BRICS_New_Delhi_Declaration_Building_for_Resilience_Innovation_Cooperation_and_Sustainability_September_12_2026

[12] https://www.nytimes.com/2026/09/14/world/asia/china-ai-security-risks-anthropic.html