Entire populations could one day be cut off from communication, clean water, heating, and financial services ‘at the click of a button,’ the United Nations human rights chief has warned, as increasingly powerful AI systems raise concerns over the safety of critical infrastructure.

Volker Türk issued the warning at an AI and Human Rights Summit in Oxford on 5 October, saying the world was running out of time to put effective safeguards in place as artificial intelligence advances at an unprecedented pace.

AI Risks Could Reach Essential Services
Türk warned that the dangers posed by AI may not always emerge through obvious failures. Instead, he said they could gradually undermine equality, freedom, privacy, and public debate, describing the risk as ‘letting dystopia creep in unnoticed.’

The warning extends beyond individual users to critical infrastructure. In a letter sent to governments and frontier AI companies in September, Türk said failures involving powerful AI systems could bring essential services to a standstill. He called for mandatory human rights safeguards and due diligence throughout the development and deployment of AI.

Countries that host AI companies or have influence over their supply chains should ensure companies act responsibly and face accountability when they fail to do so, he said. Türk also called for stronger cooperation on AI security and independent monitoring. Existing laws covering areas such as liability, cybersecurity, and data protection could provide some protection, although he said new legal approaches may also be needed.

AI Could Change Warfare and Education
Türk also warned about the use of AI in warfare, calling for urgent action to prevent machines from determining who lives and who dies. He said AI was already changing the nature of warfare and argued that decisions involving human life should remain under meaningful human control. The UN rights chief also raised concerns about AI’s impact on education.

He cited studies suggesting that students who rely heavily on AI can retain less information and develop weaker independent thinking skills. That raised a broader question for Türk: whether society could reach a point where people are no longer able to understand, challenge or meaningfully direct increasingly powerful technologies.

UN Efforts and Voluntary AI Rules

Türk pointed to recent international efforts to strengthen AI governance, including the Global Dialogue on AI Governance and the Independent International Scientific Panel on AI. He also highlighted a joint call for tighter controls on frontier AI models backed by 29 countries and the European Union.

His comments came after new commitments from AI companies and governments aimed at improving safeguards and risk assessment. A voluntary agreement signed at the White House on 29 September by US President Donald Trump and major AI executives established voluntary safety standards and included commitments to work with independent auditors.

Read More: UN Chief Warns AI Could Cut Entire Populations Off From Water, Heating and Money ‘At A Click’