User Criticizes Hiding Reasoning Traces as Contradicting AI Safety Goals

MoonL88537 · x · 2026-08-28

A user sharply criticizes Anthropic and OpenAI for hiding model reasoning traces or Chain of Thought. The user argues that if companies are serious about AI safety, they should not conceal these processes but instead open them for public scrutiny. Keeping them closed is viewed as prioritizing investor interests over the broader imperative of AI safety.

Original post →

More from Safety

Safety channel →