The United Nations has opened its first Global Dialogue on Artificial Intelligence Governance in Geneva, with Secretary-General António Guterres warning that AI is being deployed faster than societies can control it. A new UN report outlines major risks including concentration of power, mass manipulation, threats to children, and military use. 'We may be the last generation able to set the terms for coexistence between humanity and machines,’ Guterres said.
Guterres opened the three-day event on July 6, 2026, describing the current situation as an experiment conducted on societies without a plan or consent. The preliminary UN report titled 'Assessment of Opportunities, Risks and Impact of Artificial Intelligence’ serves as the basis for discussions among member states, experts, and private sector representatives.
Yoshua Bengio, co-chair of the UN’s Independent International Scientific Panel on AI, stated that science cannot currently guarantee that as AI capabilities grow, it will not cause catastrophic harm—either on its own or through malicious use. The panel includes 40 scientists and experts from around the world, among them Piotr Sankowski, director of Poland’s IDEAS research institute.
Concentration of AI power threatens sovereignty
A central theme of the report is the overwhelming concentration of AI capabilities in a few countries and companies. According to 2025 estimates, the United States accounts for 75 percent of computing power among the world’s top 500 AI supercomputers, while China holds 15 percent. The rest of the world shares the remaining 10 percent.
– The majority of states have no influence over this technology. They cannot even verify how the AI models used by their citizens operate – said Jakub Szymik, founder of the Digital Democracy Observatory Foundation, who participates in the Geneva dialogue.
Guterres proposed the creation of a Global AI Fund to help member states build skills, data resources, and access affordable computing power. Over 20 countries have already responded by nominating centers for a UN-supported Global Network for AI Capacity Building. 'We cannot allow digital exclusion to become AI exclusion, and AI exclusion to become a development gap, a security gap, and a sovereignty gap,’ Guterres stressed.
Election interference and deepfakes: The new frontier
The UN report highlights how AI is transforming information integrity and democracy. It enables mass production of persuasive content and individualized real-time manipulation that exploits users’ cognitive vulnerabilities. One cited experiment showed that an AI model optimized for persuasion shifted opposition voters’ preferences by 25 percentage points.
– The threat of large-scale manipulation resonated very strongly. We heard that AI can be used to precisely manipulate voters and ultimately tip the scales in elections – Jakub Szymik noted.
A separate report by the Digital Democracy Observatory Foundation, published on July 6, argues that voluntary commitments by tech companies—such as the 2024 Munich Declaration—are insufficient. In 2024, AI-related incidents occurred in 80 percent of countries holding national elections, with 69 percent deemed harmful. The foundation calls for mandatory labeling of AI-generated content, audits of recommendation algorithms, and a ban on deepfakes impersonating candidates.
Other risks highlighted in the UN report include the use of AI to generate child sexual abuse material (CSAM), AI-driven sycophancy that reinforces users’ beliefs—linked to several mental health incidents and documented deaths—and the deployment of AI by governments and security services for surveillance and military purposes.
Guterres concluded with a stark warning: 'The doors are still open, but they will not remain open long.’ The UN plans to present recommendations on the Global AI Fund to the General Assembly in the coming months.
Źródło: WNP.PL, Fot. UN Photo/Evan Schneider, Frame Stock Footage/Shutterstock
