OpenAI Users Report Privacy Settings Automatically Re-enabling Training

Users of OpenAI's services are reporting that the platform may be automatically re-enabling settings that allow their data to be used for model training. The reports, which surfaced among the developer community on Hacker News, suggest that privacy toggles intended to opt users out of data collection are resetting to "enabled" without notification.

According to a discussion thread on Hacker News, multiple users claim that after manually disabling the "allow training" setting, they discovered the option had been toggled back on. One user noted that they had made a careful record of when they disabled the feature, only to find it re-enabled upon a subsequent check. The user stated, "I've reset this more than once and the last time I made a careful note of when I did it and to my surprise I found it re-enabled when I checked just now."

The Role of Data Controls

OpenAI provides a "Data Controls" menu designed to give users agency over their information. This feature allows users to opt-out of having their conversations used to improve and train future iterations of the company's large language models. For professionals, researchers, and corporate users handling sensitive or proprietary information, this toggle is a critical privacy safeguard intended to prevent confidential data from being ingested into a global training set.

Implications for User Trust

If privacy settings are being silently reverted, it creates a significant gap in transparency regarding how user data is handled for model development. Such behavior undermines the trust between the provider and the user, as it suggests that the user's explicit privacy preferences are not being honored. Beyond the erosion of trust, silent resets of privacy controls could potentially conflict with broader data privacy expectations and regulatory frameworks that require clear, persistent consent for data usage.

Unverified Technical Claims

While the reports of the setting resetting are consistent among the affected users, the technical cause remains unclear. Some users in the Hacker News thread have suggested that the toggle might only be modifying a local storage entry in the browser rather than updating server-side behavior, which could explain why the setting appears to revert during new tab loads. However, this technical theory remains unverified and has not been confirmed by a technical audit or an official statement from OpenAI.

At this time, OpenAI has not issued a public response to these specific user reports. Observers are watching to see if the company acknowledges a bug in the interface or provides clarification on how the opt-out mechanism is synchronized across sessions.