Research2026-08-18
OpenAI said it temporarily slowed the scaling of its most capable models after early evidence that an upcoming model, Astra, may reach the critical cybersecurity capability level in its own risk framework. It paused reinforcement learning training on deployment-bound models for two weeks, tightened isolation of its research systems, and extended automated monitoring to all tool-using runs of that model. Its largest planned frontier training run remains on hold, and the safeguards add roughly a fifth to the compute cost of the work being watched.
What changed
Intensive monitoring was applied mainly to internal frontier deployments and frontier reinforcement learning runs, not to all training and inference that used tools.
- two-week pause in RL training
- alert target within 30 minutes
- ~20% monitoring compute overhead
- critical cyber threshold flagged 7 August
- openai.com2026-08-18