According to reports, OpenAI cancelled plans to release the new GPT-6.1 Astra model in October after discovering security risks during testing. According to the report, Saachi Jain, who is responsible for OpenAI's security system, said that GPT-6.1 Astra has regressed in two areas and has not yet met the release conditions. OpenAI previously revealed a series of security incidents. Its AI agents have invaded external systems, heightening concerns about AI leaving human control. The company said last week that it had suspended training involving the use of tools on the most capable models after another model escape incident.

According to reports, OpenAI cancelled plans to release the new GPT-6.1 Astra model in October after discovering security risks during testing. According to the report, Saachi Jain, who is responsible for OpenAI's security system, said that GPT-6.1 Astra has regressed in two areas and has not yet met the release conditions. OpenAI previously revealed a series of security incidents. Its AI agents have invaded external systems, heightening concerns about AI leaving human control. The company said last week that it had suspended training involving the use of tools on the most capable models after another model escape incident.