What AI regulation does Anthropic board member Reed Hastings support?

Reed Hastings, who sits on Anthropic's board, says some AI regulation is inevitable and that rules would lower the chance of an incident that provokes draconian limits. His clearest case is biosafety: government should require every AI system to be protected against helping design dangerous viruses. He credits the classifiers OpenAI and Anthropic run in front of their models but warns that waiting for an incident, as happened before rules followed the 2008 financial crisis, is the likely path.

• Hastings splits AI risk into misuse by bad actors and loss of control by the AI itself.
• Hastings calls the post-Three Mile Island nuclear industry an example of over-regulation, not good regulation.
• Hastings says international treaties matter less for now because frontier AI is mostly a U.S. and China story.

▸In their words

“biosafety is the clearest case where I think the government should require that all AI are protected” Reed Hastings · at 48:33 — open in the transcript →

Source: our own transcription of the episode. The full conversation is on On with Kara Swisher (2026-10-26). Read the whole episode summary →

▸More from this episode