UN High Commissioner for Human Rights Volker Turk warned on Monday that artificial intelligence could pose an "existential" risk to humanity, and called for an all-out effort to put "cast-iron guarantees" in place around the safety and security of AI before it is too late.

The warning came in a wide-ranging address to the 47-member Human Rights Council in Geneva, Turk's first since winning an unprecedented second four-year term in July, in the face of opposition from Moscow and Washington. He pledged to press AI firms to reduce risks that his office said included disruptions to services, communications and democratic systems.

"I share the concerns of industry insiders that advanced AI could pose an existential risk to humanity," Turk told the council. "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."

Turk did not elaborate on the specific nature of the risks. A spokesperson for his office said failures involving powerful AI systems could disrupt critical infrastructure and democratic institutions. According to the spokesperson's account, July's so-called Hugging Face incident — which the office described as involving a swarm of OpenAI agents hacking an open-source platform — pointed to "dangerous agent-training behaviours" and suggested frontier AI capabilities are advancing faster than the safeguards around them. The details of that incident could not be independently confirmed from the sources cited here.

Turk also warned that just a handful of men hold "almost unlimited power over AI," without naming individuals. The largest AI companies include Meta, Anthropic, OpenAI and Google. "At a minimum, we need countries hosting AI and those involved in its supply chains to come together around agreed red lines," he said.

Autonomous weapons and the push for binding rules

Turk voiced "horror" at reports that Russia had used fully autonomous drones in Ukraine, allegedly killing three people in August. Those reports have not been independently verified, and Russia's diplomatic mission did not immediately respond to a request for comment.

"I join calls for the urgent prohibition of weapons that can take lives without human involvement," he said, referring to ongoing talks in Geneva.

The warning lands two weeks after UN Secretary-General António Guterres and International Committee of the Red Cross president Mirjana Spoljaric Egger renewed their joint appeal on 25 August for a legally binding treaty on autonomous weapons. That appeal calls for prohibiting two categories: weapons whose actions are unpredictable, and weapons designed to target human beings. "We are now dangerously close to crossing a moral red line: the autonomous targeting of humans by machines," Guterres said.

The UN and ICRC first issued their joint appeal in 2023, urging states to negotiate a legally binding instrument by 2026. They are now pressing governments to move beyond discussion and begin negotiations, pointing to the seventh Review Conference of the Convention on Certain Conventional Weapons, due in Geneva in November, as the clearest path forward.

A wider agenda

The address ranged well beyond AI. Turk called for accountability for around 23 deaths he said had taken place in US immigration custody so far this year, and said he was concerned about plans to equip agents with electric shock gloves. The US Department of Homeland Security has previously said it is committed to a "safe, secure and humane" environment in detention, and its inspector general is examining deaths in custody since October 2021.

He also condemned Russia for its intensified attacks on Ukraine and Iran for rising numbers of executions. With more than 60 wars now involving at least one state — the highest since World War Two — Turk pledged greater scrutiny of the businesses and economic interests fuelling conflicts. "We will not shy away from exposing the shadow players and calling for accountability," he said.

Later this month, his office is due to publish a list of businesses operating in Israeli West Bank settlements that the UN deems illegal, a charge Israel dismisses. The council meets again on 7 October to discuss the Iran war, Sudan's civil conflict and other crises. The Trump administration disengaged the US from the council last year, accusing it of anti-Israeli bias.

Why it matters

Turk's intervention adds the UN's top human rights official to a growing institutional push — now spanning the secretary-general, the Red Cross and disarmament officials — to convert years of discussion on AI and autonomous weapons into binding commitments. The concrete near-term test is the November CCW review conference, where states will decide whether to open treaty negotiations. For AI companies, the speech signals that a UN office with a fresh four-year mandate intends to press them directly on safeguards, with the Hugging Face incident now cited as evidence that capabilities are outpacing controls.

Our take

The notable shift here is rhetorical: "existential risk" language, once associated mainly with AI lab researchers and safety advocates, is now coming from the UN's human rights machinery, and it is being paired with a specific enforcement posture rather than a general expression of concern. What remains unproven is leverage. The Human Rights Council has no direct regulatory power over AI firms, and with the US disengaged from the body, Turk's "cast-iron guarantees" will depend on whether host countries and supply-chain states translate his red-lines framing into actual negotiations — starting in November.

Sources