OpenAI to train models on Daybreak Blue cybersecurity session data

OpenAI will start using data from Daybreak Blue cybersecurity sessions to train its models on November 29, even when users have opted out of training; after backlash, a follow-up email promised a separate opt-out, reviving concerns over data use without consent.

• AI smart glasses can answer 80% of SAT questions, study finds2026-11-01
• OpenAI internal model produced 722 math manuscripts2026-10-31
• OpenAI cuts funding target by $10B; valuation falls to $1.25T2026-10-31
• GPT-5.6 Sol beats Factorio on peaceful mode2026-10-31
• Researchers hijacked ChatGPT accounts via shared links; OpenAI fixed it2026-10-31