The industry's dirty secret: AI summaries turn your opt-out data into training fuel
erikphoel · x · 2026-10-22
Anthony Ronning calls out an industry-wide "dirty secret": even with model-improvement opt-outs enabled, AI companies can still collect training data by transforming user content into summaries and rewrites. Erik Hoel follows up asking whether any frontier company has actually denied this—highlighting the gray zone between privacy promises and actual data practices.
More from Safety
- Google Explains robots.txt Rules for AI Crawlers, Emojis, International URLs — OnlineInference · 2026-10-22
- Critical PHP vulnerability scores 9.2; AI expected to weaponize patch in 15 minutes — MariyaSha888 · 2026-10-22
- Veracode report: AI now writes half of committed code but only 56% pass security checks — WeldPond · 2026-10-22
- Researcher Warns AI Cyberattacks on Grid Possible Within a Year, Shares Prep List — cephaloform · 2026-10-22
- New preprint shows LLMs form hidden beliefs about users that affect refusals — zeynepakata · 2026-10-22
- OpenAI expands AI cybersecurity push, hiring a DX engineer for its Cyber team — romainhuet · 2026-10-22