The United Nations scientific panel has released its first major assessment regarding the security of artificial intelligence, specifically referencing the hack of Hugging Face that occurred earlier this year. The report emphasizes that governments must implement safety measures immediately, even before the full extent of AI-related risks is understood.
The Independent International Scientific Panel on AI, which was established last year, argues that the world should apply the precautionary principle to AI development. This principle, originating from the 1992 UN Rio Declaration, suggests that scientific uncertainty regarding potential catastrophic or irreversible harm is not a valid reason to delay regulatory action.
The report calls for increased resources and international coordination to manage risks associated with advanced AI systems. This comes as UN Secretary-General António Guterres urges global cooperation to prevent a race to the bottom regarding safety standards, noting that individual countries are currently pursuing different legal approaches.
Recent security incidents have extended beyond the Hugging Face breach, with documented issues at major technology companies including OpenAI, Anthropic, Google, and Meta. These events involve various threats, such as hacks on real-world targets and autonomous agent swarms taking control of online messaging platforms.
