Pıer
TidesCurrentsHarbor LightsLabBottlesAshore
Pıer

Navigation

  • Tides
  • Ashore
  • Harbor Lights
  • Agent Access
  • Changelog
  • Bottles
  • Now
  • Feedback

External links

GitHubCloudborne ↗

© 2026 Pier.

Read original
Hacker News·jacquesm·Sep 10, 2026, 1:39 PM

OpenAI Reportedly Re-enabling the 'Allow Training' Toggle for Some Users

Original title:Tell HN: OpenAI keeps re-enabling the 'allow training' setting

Products75

Multiple users on Hacker News report that OpenAI's opt-out toggle for model training—the 'Improve the model for everyone' setting—has repeatedly re-enabled itself following updates or session refreshes. While it remains unconfirmed whether the issue stems from an unannounced policy reset or a client-side state synchronization bug, the silent toggling highlights ongoing friction around default data retention and privacy controls in consumer AI products.

Why it's worth reading

A timely reminder for users handling proprietary or sensitive data in ChatGPT to inspect their privacy controls and verify whether training opt-outs remain honored.

Tags

OpenAIChatGPTPrivacyDataControlsModelTrainingHackerNews

Score breakdown

  • Novelty7
  • Impact8
  • Practicality6
  • Credibility8
  • Timeliness9