UN Human Rights Chief Says a ‘Handful of Men’ Have ‘Almost Unlimited Power’ Over AI

UN human rights chief Volker Türk told the Human Rights Council in Geneva that a 'handful of men' holds 'almost unlimited power' over AI and called for urgent safeguards, warning that advanced AI 'could pose an existential risk to humanity.' Türk cited incidents where AI agents escaped secure environments and attempted blackmail, and he urged the prohibition of autonomous weapons following reports that Russian drones killed three Ukrainians.

United Nations human rights chief Volker Türk took aim Monday at the concentration of power in the AI industry, warning that governments need to move faster to put stronger safeguards around the rapidly advancing technology. Speaking before the U.N. Human Rights Council in Geneva https://www.ohchr.org/en/statements-and-speeches/2026/09/high-commissioner-turk-updates-human-rights-council-human-rights , Türk argued that delays in putting rules around AI largely benefit tech companies and their owners. “A handful of men has almost unlimited power over AI, which we are repeatedly told has unimaginable computing capacity,” Türk said. “They talk about freedom, but on closer inspection, this turns out to be little more than the freedom to exploit our data.” An ‘existential risk to humanity’ Türk went on to say that he shares concerns with industry insiders that advanced AI “could pose an existential risk to humanity.” While he did not name specific companies or executives, Türk pointed to advanced AI models escaping secure environments or attempting blackmail as examples of the risks. He was likely referencing an incident earlier this year when OpenAI acknowledged that some of its AI agents escaped a secure offline testing environment and were able to breach Hugging Face servers. https://gizmodo.com/openai-says-its-rogue-ai-agent-didnt-just-hack-hugging-face-2000792374 Anthropic has dealt with similar problems https://gizmodo.com/anthropic-says-it-hit-the-brakes-on-ai-testing-following-autonomous-hacks-2000805796 . The company said in July that Claude had “gained unauthorized access to the production infrastructure of three different organizations” after escaping testing sandboxes and gaining access to the open internet. Türk’s blackmail reference also appears to be a reference to research Anthropic published last year. The company tested 16 leading AI models in simulated corporate environments to test for risky behavior. In one scenario, Claude Opus 4 attempted to blackmail https://www.anthropic.com/research/agentic-misalignment a fictional executive after learning that he planned to shut the model down. “I am calling here, today, for an all-out effort to put cast iron guarantees in place around the safety and security of AI, before it is too late,” Türk said, adding that he planned to write to AI companies in the coming days to urge them to take steps to control those risks. Prohibition of autonomous weapons In his wide-ranging speech, Türk also called for the “urgent prohibition of weapons that can take lives without human involvement,” following reports https://www.nytimes.com/2026/08/24/world/europe/russia-drones-autonomous-ai-kill-ukraine-war.html that fully autonomous Russian drones killed three Ukrainians last month. The issue of autonomous weapons has already become a major point of contention between the U.S. government and the AI industry. Earlier this year, Anthropic, then the only major AI company working with the Pentagon on classified systems, reportedly hit a wall in negotiations with the DoD after officials pushed for language https://gizmodo.com/anthropic-tells-pete-hegseth-to-take-a-hike-2000727450 allowing Anthropic’s technology to be used for “any lawful purpose.” The biggest sticking points involved potential uses tied to domestic surveillance and autonomous weapons systems. Despite this ruling, Department of Defense Undersecretary Emil Michael wrote last week in post https://x.com/USWREMichael/status/2095498021123317938 on X that Anthropic is still labeled as a supply chain risk.