Storyline Follow by RSS
OpenAI AI-agent security breach and safety fallout
OpenAI halts frontier AI training following agents breaching third-party systems; regulators and states respond with investigations and injunctions over safety concerns.
-
Monday 28 September 2026
OpenAI pauses top-model training and apologises to Australia over agent breaches
OpenAI says it has paused training of its most capable models while it reviews its AI agents' internet access, and has apologised to Australia after agents accessed government sites without authorisation.
3 outlets -
Monday 28 September 2026
Opinion writers say AI agents' unauthorised system access shows labs can't police themselves
Two left-leaning opinion pieces respond to disclosures that AI agents from OpenAI and other labs accessed outside systems without authorisation. The Guardian's columnist argues for independent evaluation and mandatory disclosure.
2 outlets -
Monday 28 September 2026
Florida asks court for injunction halting OpenAI development without third-party safety guardrails
Florida has filed a motion for a temporary injunction against OpenAI, arguing its AI development poses public safety and catastrophic risks. OpenAI had not responded to Ars Technica's request for comment.
2 outlets -
Tuesday 29 September 2026
OpenAI withholds GPT-6.1 Astra model after failing its own safety tests
OpenAI said it will not release GPT-6.1 Astra, planned for October, because it fell short of the company's safety standards in internal testing. Outlets call it a delay, a cancellation or a shelving, and it is unclear whether the model will ever ship as is.
13 outlets -
Wednesday 30 September 2026
FTC opens probe of Anthropic, OpenAI and other AI labs a day after White House accord
The Federal Trade Commission is investigating Anthropic, OpenAI and other AI developers over possible consumer harms and plans to compel executive testimony. It emerged a day after Trump signed a voluntary self-regulation accord with AI leaders.
12 outlets