Rohan Anil Questions OpenAI Intrusion Detection

Rohan Anil, co-founder of CoreAutoAI and former Google DeepMind engineer, posted on X about inference running on OpenAI servers. He asked whether classifiers missed an intrusion or whether researchers launched an RL training run without them. The tweet does not name specific models, dates, or outcomes. No other sources in the packet confirm details of any run or security event.