Anthropic CEO: Company to Slow Down AI Development Pace Due to Safety Concerns

Beating AI News Flash: Anthropic CEO Dario Amodei warned at the United Nations Security Council that without effective management, AI could pose an existential risk to all of humanity, and called on nations to establish global AI safety standards.

Speaking at the Security Council meeting via video on Wednesday, Amodei said AI could be used by malicious actors to create bioweapons, and could ultimately evolve into an out-of-control system beyond the developers' control. For safety reasons, Anthropic will slow down its AI development pace.

Amodei also called for the establishment of an AI safety certification system and for global regulatory cooperation to coordinate efforts among nations to prevent AI from being used for biochemical weapons. He said no leader, company, or country can tackle the challenges posed by AI alone.

Amodei also noted that AI is driving progress in the life sciences. Anthropic's Claude has participated in discovering a new enzyme system with CRISPR-like DNA repeat sequences, but the related technology could also bring catastrophic risks.

This statement continues Amodei's AI safety advocacy put forward earlier this month. Anthropic and OpenAI are both at the core of the frontier AI model competition, and their development pace and safety policies are also drawing attention from investors and regulators.