OpenAI plans to use Daybreak Blue session data to train models
The story so far
OpenAI will begin using previously untrained Daybreak Blue session data to train models starting November 29, raising concerns among researchers about incidents similar to Buckmaster-Alpöge. OpenAI confirmed this policy change to Trusted Access members and committed to providing a separate opt-out option, but interactions flagged during security audits can still be used for security purposes. This move is based on OpenAI's statement that interactions flagged as security issues can continue to be used for security audits. The plan does not mention whether there will be a choice for the data usage of all participants, only confirming the provision of an opt-out option.
AI-written from the reports · Updated 18 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- The DecoderOpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits
OpenAI will start using previously untrained Daybreak Blue session data to train models on November 29, raising concerns among researchers about incidents similar to Buckmaster-Alpöge. OpenAI confirmed this policy change to Trusted Access members and committed to providing a separate opt-out option, but interactions marked during security reviews can still be used for security purposes.