UN Warns Military AI Threatens Humanity, Calls for Binding Controls

Global leaders and experts are raising urgent concerns about the risks posed by artificial intelligence, particularly in military and security contexts. The Global Commission on the Responsible Use of AI in the Military has urged legally binding agreements to keep nuclear weapon decisions under human control, warning that AI's speed and accuracy could lead to catastrophic outcomes if unchecked. Experts like Microsoft's AI CEO Mustafa Suleyman and Google's DeepMind CEO Demis Hassabis have highlighted the potential for AI to become so powerful it might require military-grade intervention and noted society is not yet prepared for advanced AI capabilities. Beyond military uses, AI's expansion in businesses raises significant cybersecurity threats, including data poisoning and manipulation of AI models, which could compromise sensitive systems and facilitate cyberattacks. The UN's Special Envoy Lutfe Siddiqi emphasized both AI's benefits for peacekeeping and humanitarian efforts and its dangers, such as autonomous weapons, disinformation campaigns, and algorithmic bias, urging international cooperation to mitigate these risks. Together, these perspectives underscore the pressing need for global governance and coordinated policies to harness AI's advantages while preventing its misuse and ensuring human oversight in critical decisions.