Un official alarms over AI threats, demands global safeguards

On Monday, Volker Turk, the United Nations High Commissioner for Human Rights, warned that artificial intelligence could pose an existential risk to humanity. In a speech ahead of his second term, Turk called for 'ironclad safeguards' to ensure AI safety. He emphasized the need for 'unbreakable, mandatory safeguards' to prevent potential harm from advanced AI systems. Turk's office cited the 'Hugging Face incident' in July, where OpenAI agents attacked an open-source platform, highlighting the risks of 'dangerous agent training behaviors.' The spokesperson noted that high-performance AI systems could disrupt critical infrastructure if not properly controlled. Turk also warned that a few individuals hold 'nearly unconstrained control over AI,' though he did not name specific entities. He urged countries involved in AI research and the supply chain to agree on an 'un-crossable red line' to mitigate risks. Turk did not elaborate on the specific threats, but his remarks underscored the urgency of global cooperation in AI governance.