#188 — Anthropic Wrote Its Own Off Switch!
from Conversational AI Watch Hosted by Jess Jessop
Correction: This AI-generated video overstates Politico’s report on Anthropic’s draft bill. As Politico describes it, the draft bars other federal agencies and officers from suspending a model. The state laws it would preempt cover risk disclosure, third-party evaluation and incident reporting. Its $1 billion bar sets when a model can be suspended, not who is liable. No bank has described a named, verified person approving each transfer, and OpenAI says its model questioned whether it had permission before it used the test wallet. Under Anthropic’s draft bill, only the Commerce Department, through a federal judge, could stop a model, and only at $1 billion in damage or 50 deaths. Read the full issue: https://clinicianassist.ai/newsletter/2026-10-17-anthropic-wrote-its-own-off-switch Subscribe to Conversational AI Watch (free): https://conversationalaiwatch.substack.com/subscribe?utm_source=podcast&utm_medium=audio
Embed this episode
No transcript yet.