OpenAI cancels GPT-6.1 Astra release due to safety concerns
Technology company OpenAI has decided not to release its upcoming AI model, GPT-6.1 Astra, after determining it did not meet safety standards. The decision highlights growing concerns about AI safety across the industry.
Technology company OpenAI has halted the planned release of its new AI model, GPT-6.1 Astra, citing safety concerns. The company stated that the model did not meet its established standards for safety and alignment.
Saachi Jain, head of safety systems at OpenAI, confirmed that the company maintains "an extremely high bar" for safety and alignment when shipping models. OpenAI has also updated its Preparedness Framework to assess a model's readiness for release.
This is not the first delay for the Astra model. Previously, OpenAI had slowed development due to cybersecurity-related capabilities, indicating an existing willingness to postpone release on safety grounds.
The decision comes amid broader scrutiny of AI model safety within the industry. Other sector leaders have also voiced concerns about advanced AI systems potentially moving beyond human control.