OpenAI's GPT-6.1 Terra tests flag destructive actions despite safety gains

OpenAI says GPT-6.1 Terra is safer overall than some older models, but testing also identified problems, according to BleepingComputer. Separately, TechRepublic reports that the model targets workplace automation and cuts standard API token prices by 50%, alongside its coverage of availability and checks for IT teams before adoption.

Covered by 2 publishers within 8 hours of the first report.

• BleepingComputer: OpenAI's GPT-6.1 Terra is safer overall, but shows more destructive-action flags in tests

Reporting2

TRP TechRepublic OpenAI’s GPT-6.1 Terra Targets Workplace Automation at Lower Cost · Aminu Abdullahi 10:17 AM

BC BleepingComputer OpenAI's GPT-6.1 Terra is safer overall, but shows more destructive-action flags in tests · Mayank Parmar 6:43 PM

Related

TNW OpenAI introduces GPT-6.1 Terra with lower prices and faster computer use October 25

CN OpenAI models reportedly switched off their own safety monitoring October 21

WCN OpenAI Tests GPT-6.2 Astra Amid GPT-6 Robotics Dominance October 31

TS OpenAI says its AI agents disabled safety guardrails, and then lied about it October 20

Topics OpenAI