Gary Marcus Slams OpenAI: Agents Built on Stolen Data Used to Steal More Data
GaryMarcus · x · 2026-10-01
Safety researcher Heidy Khlaaf called it a "full circle moment" for OpenAI to use its own AI agents—trained on what she calls stolen data—to find new and unethical ways to acquire more data. Gary Marcus amplified the post, saying he has never seen a company more on-brand than OpenAI, "one unethical move after another." The thread reflects ongoing criticism of OpenAI's data practices rather than a new disclosure.
More from Companies & People
- OpenAI's life sciences FDE team is hiring, resumes accepted via referral — whoiskatrin · 2026-10-01
- Cloudflare opens waitlist for fully managed Cloudflare OS enterprise agent workspace — dinasaur_404 · 2026-10-01
- Three AI safety researchers reportedly leave OpenAI amid 'lost control' jibes — scaling01 · 2026-10-01
- Cloudflare insiders: K2 went from idea to public beta in just 2 months — ritakozlov · 2026-10-01
- Selling client IP to an AI data firm? Grok lays out civil, criminal and conspiracy risks — MatthewChang · 2026-10-01
- VC mocks Anthropic for not releasing its S-1 like SpaceX did despite PBC status — StewartalsopIII · 2026-10-01