Tell HN: OpenAI keeps re-enabling the 'allow training' setting
Points and comments are a snapshot, not live.
User reports OpenAI keeps re-enabling the 'allow training' setting after they turned it off.
A Hacker News user reports resetting OpenAI's 'allow training' setting multiple times. After carefully noting when they last disabled it, they found it re-enabled upon rechecking. The user warns others to verify the setting remains off.
What commenters are saying
Commenters are split: some suspect deliberate behavior by OpenAI, others blame incompetence or bugs in distributed systems. One commenter notes the toggle sets a local storage entry whose value may not matter. Another suggests documenting the change to strengthen legal claims. Several users recommend switching to competitors like Bedrock, though others criticize Anthropic's data-sharing practices too. A commenter reports their own setting was re-enabled despite being previously disabled.