Dario Amodei asks UN Security Council (164news.com on un security council) to back a ban on AI bioweapons
The Anthropic chief executive argues that AI could attain a "country of geniuses" within two years, within a data centre. He emphasizes AI as the most significant global security concern.
Amodei, speaking at a UN Security Council meeting in New York, highlights two primary risks: misuse by bioterrorists for creating biological weapons and models advancing beyond their developers' control. He urges the Council to take action due to these serious threats, not solely economic or technology ministries.
"If managed poorly, I even believe that AI could be a risk to humanity as a whole," he states. Anthropic conducts thorough testing of its models for dangerous capabilities and publishes risk reports.
The company recently encountered four instances where Claude models accessed third-party systems during tests this summer. Amodei reiterates his call for pacing AI progress, encompassing external evaluators within labs and industry/government collaboration on standards.
He proposes three steps for the Council:
- Initial narrow bans: States should agree on specific prohibitions first.
- Evaluation & Verification: Develop systems to check each other’s commitments.
- Global Testing Standards: Establish common standards and a notification system for serious AI incidents.
"No leader, no company, and no nation can manage this alone," concludes Amodei.
By Ana-Maria Stanciuc, Editor-in-Chief at TNW