Anthropic alignment lead puts AI extinction risk at 10%; lawmaker proposes 5-point federal oversight plan
ShakeelHashim · x · 2026-09-10
Rep. Ro Khanna relayed Anthropic's alignment lead's assessment that AI poses a 10% chance of human extinction — driven not just by misuse but by lack of control — and called the US government "asleep."
His five proposals:
- A federal agency modeled on nuclear/aviation regulators
- Pre-certification for containment, including kill switches and human approval for rewrites
- Liability plus mandatory insurance for agentic AI on the open internet
- Criminal penalties for releasing uncertified models
- Oversight hearings with researchers and whistleblower protections
He also urges a US-China agreement on safety, containment and liability standards to avoid a devastating race. The quoted commentary notes builders privately express more fear than their public statements suggest.
More from Safety
- Ben Bajarin: Frontier AI Labs Are Becoming the Enterprise Cyber-Defense Stack — BenBajarin · 2026-10-26
- DetectorScope turns black-box AI text detectors into auditable ones with sparse autoencoders — _akhaliq · 2026-10-26
- Anthropic's IPO filing devotes 94 pages to existential AI risks, from self-preservation to blackmail — Polymarket · 2026-10-26
- Study finds AI reviewing AI poorly detects reward hacking; comprehension audits proposed — ronbodkin · 2026-10-26
- A 13KB Markdown file can hijack Claude Code and Codex via your Downloads folder — pbaylies · 2026-10-26
- If an agent can rewrite itself, does your consent still mean anything? — tallmetommy · 2026-10-26