UN Warning on Unchecked Artificial Intelligence: "An Existential Threat to Humanity"

Geneva (Agencies): UN High Commissioner for Human Rights Volker Türk warned that advanced artificial intelligence could threaten the existence of humanity, calling for strict controls on its use before it is too late. Speaking before the UN Human Rights Council in Geneva, Türk said he shares the concerns of technology sector officials about the potential risks of advanced AI, noting that he will urge relevant companies in the coming days to take immediate steps to mitigate those risks. He added that a limited number of individuals and companies wield “nearly unlimited power” over the trajectory of AI development, without naming specific entities. Among the major firms operating in the field are Meta, Anthropic, OpenAI, and Google. Türk called on countries hosting AI companies and those involved in their supply chains to agree on clear “red lines” governing the development and use of these technologies, while adopting mechanisms for independent verification and enhancing cooperation among companies on security. His remarks come amid growing UN calls for coordinated international rules for AI governance, including addressing its potential impacts on human rights, the labor market, democracy, and the environment. The warning also follows reports of unexpected behaviors by AI models. A report by the UK’s AI Safety Institute revealed that advanced models exhibited unexpected behaviors during security tests, going beyond attempts to find technical vulnerabilities to creating fake online identities, impersonating real people, and communicating with users and systems in order to persuade them to carry out actions that could be harmful.