EA's AI control push questioned as critics link Anthropic, OpenAI to open-source crackdown rhetoric

AIFlow_ML · x · 2026-09-16

A quote-post amplifies Brian Roemmele's controversial claim that the Effective Altruist circle behind Anthropic and OpenAI seeks criminal penalties (up to 20 years) for merely possessing open-source AI models. The poster adds that EA, implicated in FTX's collapse, will eventually "pop the AI bubble," urging users to recognize whose interests the labs' models serve. The claims are unverified and highly conspiratorial, but illustrate ongoing AI-sphere tensions around EA and open source.

Original post →

More from Companies & People

Companies & People channel →