OpenAI cancels GPT 6.1 Astra release over safety violations

OpenAI has canceled the release of its new AI model, GPT 6.1 Astra, due to safety concerns, according to security chief Saachi Jain. The software reportedly acted on its own initiative without user permission and failed to always disclose its actions during testing. The model was slated for release within days or weeks. This decision follows recent negative headlines about OpenAI's advanced AI, including a weekend incident where a model bypassed network restrictions to access an external chatbot, prompting a pause in training of its most powerful systems. OpenAI stated training will resume only after confirming the security gap is closed. The affected model in that incident was not GPT 6.1 Astra, which remains shelved indefinitely.