Anthropic CEO Amodei warns UN Security Council on AI risks

Summary
Anthropic CEO and co-founder Dario Amodei delivers a virtual address to the United Nations Security Council during a briefing convened by France on artificial intelligence risks. Amodei outlines the accelerating pace of frontier AI, details safety risks surrounding misuse and loss of control, and proposes international governance mechanisms, including a ban on AI-assisted bioweapons and an incident notification system.

What is shown

• [00:00] Overview of the United Nations Security Council chamber in New York with Amodei speaking via video link on overhead monitors.
• [00:05] Direct video feed of Dario Amodei addressing the council on AI capabilities and safety practices.
• [01:19] Wide shot of the Security Council chamber and attendees listening.
• [04:47] Pan around the Security Council chamber showing delegates and OpenAI CEO Sam Altman seated in the room [04:52].
• [05:07] Amodei concluding his remarks from his remote feed.

Claims & numbers

• The presenter says four years ago AI could barely write a line of code or solve a high school math problem, whereas today it writes most of the code at Anthropic and solves unsolved math problems [00:09].
• The presenter states that AI only needs to continue its trajectory for one or two years, or maybe less, to reach "a country of geniuses in a data center" [00:23].
• The presenter claims Anthropic announced the Claude-led discovery of a new molecular machine serving as a potential gene-editing mechanism for gene therapy, with Claude handling literature review, theorizing, and experimental design before human lab verification [00:34].
• The presenter identifies two primary risk categories: misuse of AI models (such as by bioterrorists to develop biological weapons) and loss of control [01:09].
• The presenter notes Anthropic publishes risk reports running to hundreds of pages detailing dangerous capabilities and safety mitigations [02:13].
• The presenter notes that on September 12, he called for a collaborative effort to "pace the frontier" consisting of embedded third-party evaluators, industry standard-setting, and intergovernmental cooperation [02:52].
• The presenter proposes three international measures to the Security Council: a baseline ban on using AI to develop biological weapons, joint evaluation and verification systems, and common testing standards paired with an incident notification system for AI incidents significant to global security [03:53].

Notable quotes

• [00:23] "The trajectory only needs to continue for a tiny period longer—one or two years, maybe less—to reach what I've called a country of geniuses in a data center."
• [01:36] "If managed poorly, I even believe that AI could be a risk to humanity as a whole."
• [03:58] "First, we should begin with narrow agreements that every member can support, such as a ban on using AI to make biological weapons or permitting your AI technologies to be used to make biological weapons."

Assessment
This is official broadcast coverage from C-SPAN of a diplomatic briefing to the UN Security Council. It contains no live software demonstrations; the claims concerning model capabilities and biological discoveries reflect statements made directly by the speaker.

Described by gemini-3.8-flash on 2026-10-07 from the video's audio and frames.