OpenAI's ChatGPT Training Opt-Out Slammed as Buried and Ineffective

From September 9–10, OpenAI's training data opt-out mechanism drew concentrated criticism. Multiple users found that toggling off "Improve Model for Everyone" does not actually stop OpenAI from training on chat data; a genuine opt-out requires a buried privacy request, and its scope reportedly covers only ChatGPT, not internal models.

Confirmed

• ShashwatGoel7 and others shared that paid users can disable "Improve the model for everyone" under Settings → Data Controls, while complaining the option's naming is unintuitive.
• Screenshots from edoardocontente's thread showed "Do not train on my content" must be manually checked rather than being on by default.
• Multiple posts (RexDouglass, niloofarmire, PMinervini, xennygrimmato and others) pointed to privacy.openai.com as the real path for opting out of training data use, not the in-app toggle.
• Investor Delip Rao found that even after completing the process, the opt-out only excludes "ChatGPT," with no commitment regarding internal models such as reward models.

Unconfirmed

• paulcal countered that the in-app switch is effective and the privacy form also works, citing the Data Controls FAQ; the two sides disagree on the actual effect of each mechanism, with no official ruling.
• Delip Rao's claim about internal models not being covered is his personal interpretation; OpenAI has not publicly responded.

Why it matters

• The episode exposes usability problems in OpenAI's data controls—naming, defaults, and buried paths—meaning ordinary users may believe the toggle protects their data when the real opt-out threshold is far higher.
• If the opt-out indeed covers only ChatGPT and not reward models, users' actual control over data usage is sharply limited, directly affecting compliance decisions for enterprises and researchers.

2026-09-09 ~ 2026-09-10 · 7 related posts

• How paid ChatGPT users can turn off OpenAI training on their data — xennygrimmato_ · 2026-09-09
• [source] Debate over OpenAI data opt-out: in-app ChatGPT toggle is effective, per FAQ — paul_cal · 2026-09-09
• Researchers claim OpenAI's opt-out of training data is misleading and only covers ChatGPT — PMinervini · 2026-09-09
• Toggling off "improve mode" doesn't stop OpenAI from training on your data — RexDouglass · 2026-09-09
• OpenAI's 'Do not train on my content' option found not enabled by default — anshulkundaje · 2026-09-09
• [source] Toggling off 'Improve Model for Everyone' doesn't stop OpenAI training on your chats — PsychicorAI · 2026-09-10
• [source] OpenAI accused: buried opt-out only excludes ChatGPT, says nothing about internal models — niloofar_mire · 2026-09-10