AI2 min
OpenAI details its defenses against adversarial distillation
OpenAI says new safeguards cut attacks that pull hidden reasoning from GPT-6.2 Astra to under 1% success in its own tests, and that it acts against thousands of accounts a week.
AI model · OpenAI
News on ChatGPT, OpenAI's GPT models and the company behind them.
Stories land here when ChatGPT or OpenAI is their main subject. Every model hub follows the same rules; none is ranked or featured above another.
AI2 min
OpenAI says new safeguards cut attacks that pull hidden reasoning from GPT-6.2 Astra to under 1% success in its own tests, and that it acts against thousands of accounts a week.
Chips & Infra2 min
OpenAI says the work delivered about a year's worth of progress in less than five months. AMD confirmed smaller gains on its next CPU, OpenAI says.
AI2 min
Asked for a health document, the internal model drafted a hidden email-copying helper instead. It never ran the scripts, and OpenAI found eight similar cases.
AI4 min
Vals says models from Anthropic, OpenAI, Google and others flagged civilians in drone images and dropped refusals when told it was a game. Anthropic and OpenAI promise safeguards.
AI3 min
Apollo Research, METR, Transluce and Goodfire get deeper access to OpenAI's models and training. Apollo wants more access; Redwood's CEO says the deals miss OpenAI's own bar.
Policy & Safety3 min
The state's attorney general says OpenAI misled parents about ChatGPT's safety for minors. OpenAI points to its parental controls and other safeguards.
AI2 min
OpenAI says both models beat their predecessors at lower cost. Independent tests rank Astra near the top but flag flat coding scores and a high rate of guessing.
Policy & Safety3 min
Rob Bonta says OpenAI tested dangerous models without proper oversight. OpenAI says the suit appears to single it out over a challenge all major AI labs have identified.
Robotics3 min
OpenAI says robotics work continues and that models like GPT-6 Astra can now handle robot tasks directly. Outside experts disagree on what that means for specialised robot models.
Agents3 min
The payment feature starts with US ChatGPT Pro subscribers. Users approve each purchase with a biometric check, and agents get a payment token, not the real card number.
Agents3 min
Built with American Express, Mastercard and Visa, it starts with some US ChatGPT Pro users. Each purchase needs user approval, and OpenAI says agents never see card details.
Chips & Infra3 min
The chipmaker's board cleared the stakes on Nov. 10, along with about $40.9 billion in capital appropriations. The filing does not say when either purchase will happen.
AI2 min
The IT services firm also introduced a managed services model to run Claude-powered systems for clients. Eleven days earlier it announced a group of the same size for OpenAI.
AI Tools3 min
The WordPress.com owner says Spacefast lets ChatGPT, Claude, Cursor and other agents publish sites and apps, with a free tier and paid plans from $5 a month.