OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits

The Decoder 31 Oct 2026, 16:38 UTC 1 min read

In brief

OpenAI will begin using data from Daybreak Blue cybersecurity sessions to train its models on November 29, according to an email sent to Trusted Access program members. The company later said it would provide a separate opt-out after members objected to the change.

Daybreak Blue session data had previously been excluded from training, including when users disabled the “improve the model for everyone” setting.

Interactions identified in safety reviews may still be retained for safety-related purposes.

Researchers have raised concerns linked to the earlier Buckmaster-Alpöge episode involving Navier-Stokes drafts uploaded to Codex.

The mathematicians involved were Tristan Buckmaster and Levent Alpöge.

Why it matters: The policy change highlights ongoing questions about consent, data boundaries and researcher trust when AI systems process sensitive technical work.

Editorial brief by Sakhanda Wire AI, based on reporting by The Decoder.

OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits

Matthias Bastian View the LinkedIn Profile of Matthias Bastian

Oct 31, 2026

Nano Banana Pro prompted by THE DECODER

Starting November 29, OpenAI will use Daybreak Blue session data to train its models. An email to Trusted Access program members confirmed the change. Previously, this data was excluded from training, even when users had the "improve the model for everyone" setting turned off. Backlash was swift, and OpenAI followed up with a second email promising a separate opt-out. Interactions flagged during safety reviews can still be used for safety purposes.

OpenAI's email to Trusted Access members about the policy change for Daybreak Blue sessions.

The change has researchers worried about a repeat of the Buckmaster-Alpöge case. Mathematicians Tristan Buckmaster and Levent Alpöge uploaded Navier-Stokes equation drafts to OpenAI's Codex. OpenAI soon presented its own proof but denied using their inputs for training. But the suspicion alone was enough to show researchers what could go wrong.

AI News Without the Hype – Curated by Humans

Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.

Source: via Yoav Alon

OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits

AI News Without the Hype – Curated by Humans

More stories