OpenAI ships three safety/research posts in a day, reigniting Anthropic transparency debate

daveholtz · x · 2026-11-19

A commentary argues OpenAI shares far more useful public information than Anthropic — more misalignment disclosures, more on AI R&D automation pace and AI control — and fosters a more open culture. Screenshots show OpenAI publishing three posts on Nov 18: "Managing frontier AI risk: From training to deployment," "Research acceleration: The jagged frontier" (internal data on coding agents accelerating the AI dev cycle), and "Inside the Codex sandbox."

Original post →

More from AGI Musings

AGI Musings channel →