VQV Signal
A Hacker News post describes a company that had turned off data sharing in Codex, Claude Code and similar tools, then found that the Codex setting had been turned back on without anyone opting in. The author says the change was noticed the day before the post and points to the "Improve the model" option in ChatGPT's data controls settings. The account is a user report, and the source does not include an official statement from OpenAI.
A Hacker News post describes a company that had turned off data sharing in Codex, Claude Code and similar tools, then found that the Codex setting had been turned back on without anyone opting in. The author says the change was noticed the day before the post and points to the "Improve the model" option in ChatGPT's data controls settings. The account is a user report, and the source does not include an official statement from OpenAI.
AI-assisted summary based on the listed source.
My company is pretty paranoid about leaking IP to the frontier labs through prompts, so we've all been diligent about disabling data sharing in Codex/Claude Code/etc. But yesterday my coworkers and I noticed that the setting had been silently reenabled in Codex. None of us knowingly opted back in. You can check...
Teams that rely on opt-out settings to protect proprietary code or prompts may need to recheck their data controls for Codex and similar tools. Individual users who opted out of training should also verify the toggle, since the report suggests the setting can change without a clear notice. Watch for an official explanation from OpenAI.
VQV organizes public signals from inspectable sources. It does not independently verify the underlying report.
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
OpenAI has a source-backed update with coverage spanning for developers.
Users report OpenAI Codex data-training setting was re-enabled without their knowledge
Users report OpenAI silently re-enabled model training on Codex chats
VQV surfaced this signal because it is recent, relevant to Consumer AI, connected to Hacker News Front Page.
No login, cookies, social SDKs, or automatic posting.
