The UN Calls for Immediate Rules to Address AI Risks

Key Takeaways

1Security incidents related to AI have been reported at Hugging Face, OpenAI, Anthropic, Google, and Meta

2A scientific panel from the UN calls for the precautionary principle to be applied without waiting for definitive evidence

3The report demands more resources and international coordination for AI safety

4This call comes as the UN General Assembly and US-China discussions on AI are taking place

Why it matters : The UN is placing AI safety at the center of the diplomatic agenda and is calling for immediate measures in response to risks deemed potentially catastrophic.

A UN scientific panel has published its first thematic report and is calling for strengthened protections for AI without waiting for definitive evidence of risks. This call, based on the precautionary principle and recent incidents, comes as the General Assembly is taking place in New York and as Washington and Beijing discuss AI.

Documented Intrusions and Takeovers in Several Companies

Since the revelation of the Hugging Face hack earlier this year, incidents have been reported at OpenAI, Anthropic, Google, and Meta. These include attacks on real targets and swarms of agents taking over online messaging forums. The hack targeting Hugging Face is said to involve OpenAI.

The Precautionary Principle at the Heart of the Panel's Alert

The panel asserts that it is unnecessary to wait for the scientific community to precisely explain the origins of the incidents in order to implement stricter safeguards. It links this stance to the precautionary principle, designed for potentially catastrophic or irreversible risks even when their probability is uncertain. Enshrined in the 1992 Rio Declaration, this principle posits that uncertainty does not justify delaying action and has influenced environmental and public health policies, particularly in the European Union.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Increased Resources and International Coordination Required

Significantly greater resources are being demanded to address emerging risks associated with advanced AI systems. It also calls for enhanced international coordination on security and accountability, especially in a context where legal frameworks remain heterogeneous across countries.

A Position Statement Amid Diplomatic Engagements

This publication comes during the United Nations General Assembly in New York and as the United States and China hold discussions on AI. The previous week, António Guterres urged governments to cooperate in the face of AI-related threats, warning that the world cannot afford a race to the bottom in terms of security. The document constitutes the first thematic report from the Independent International Scientific Panel on AI, created last year as the first global scientific body dedicated to AI within the UN, and represents the organization's first major assessment of the Hugging Face hack.