Today for AI
HOT RADAR
industryHEAT 14.1°

AI CURATED HOT STORY · 10/23/2026

OpenAI Fires Three Safety Researchers Amid Internal Leak Investigation

1 reports archived1 independent sourcesupdated 10/23/2026, 14:04:55
Synthesis & Latest Updates
1 Sources Cross-Validated

OpenAI fired three internal safety researchers after security staff discovered disappearing Signal chats with an outside researcher. The action stems from an investigation into the leak of design details for the new model Astra, which sparked concerns over reduced transparency in reasoning steps.

LATEST/OpenAI fired three internal safety researchers after security staff discovered disappearing Signal chats with an outside researcher. The action stems from an investigation into the leak of design details for the new model Astra, which sparked concerns over reduced transparency in reasoning steps.

HEAT TRENDHourly heat curve

2 fully observed hours
Fewer than 3 fully observed hours; trend not plotted yet.

TIMELINECoverage timeline

Total 1 reports · Latest first
  1. Rohan Paul (@rohanpaul_ai)T2·68 pts
    • OpenAI launched an internal probe to trace the source of media leaks regarding the new model Astra's design.
    • Astra uses a method that reduces recorded reasoning steps, raising concerns about the interpretability of AI decision-making.
    • Security teams monitored employee communications and fired involved researchers, indicating extreme measures for core IP protection.