AI researcher argues OpenAI and Anthropic spend billions on safety while rivals race recklessly

Afinetheorem · x · 2026-09-09

Afinetheorem calls this an underreported story: OpenAI and Anthropic plus affiliated nonprofits spend billions on AI safety, and many high-level staff signed the Pacing letter. He argues the irresponsible actors are instead the other companies — especially non-US ones — racing to release "sovereign" and open-weights models.

Original post →

More from Safety

Safety channel →