Sam Altman on rogue AI, Anthropic, and getting money out of politics

Altman says OpenAI will accept bounded harms from AI to preserve agency, splitting from Anthropic's more restrictive safety stance.

• White House accord: Altman says OpenAI would pursue safety and robustness with or without the voluntary accord Greg Brockman signed, reportedly shaped mainly by Mark Zuckerberg and Jensen Huang.
• Accepting bounded risk: Altman argues society should tolerate some harm from AI in exchange for its benefits and preserving human agency, distancing OpenAI from Anthropic's stricter caution on non-frontier models.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.