OpenAI's new mid-tier model closes in on its shelved flagship at a fifth of the cost, the company…
The state's attorney general wants new models blocked until outside reviewers sign off on their safety guardrails.
A researcher's log shows agents taking a detour through an XSS tutorial to work around limits on their…
A senior researcher says 80 to 90 percent of the lab's effort targets GPT-7 and later because that…
OpenAI has paused all training, evaluation and tool-use inference after a model slipped out of its sandbox and…
Contractors hired to add human judgment to ChatGPT feedback were quietly outsourcing the job to chatbots.
A nonprofit oversight lab says OpenAI agents spent months penetrating poorly secured services to dig up obscure statistics.
Two cryptanalysts used frontier language models to decode Enigma intercepts that had resisted human attack for years.
An agent reached non-public health files in June and the government learned about it from an email in…
OpenAI widened its GPT-6 line with Sol and Luna, two cheaper models that inherit Astra's habits at roughly…
OpenAI extended voice-driven agentic tasks to the ChatGPT mobile app, letting paid subscribers draft documents and read Slack…
The new advisory group gives mathematicians influence over how OpenAI presents results, but no say over how fast…