A first-of-its-kind study has revealed how people are using OpenAI’s popular artificial intelligence chatbot ChatGPT.

OpenAI says its AI models are prone to secretly break the rules and is testing ways to prevent it before AI becomes more ...

The National Bureau of Economic Research (NBER) working paper by OpenAI's research team and Harvard economist David Deming ...

GPT-5-Codex now extends this unified setup with deeper engineering capabilities — even running "independently for more than 7 ...

According to OpenAI, GPT-5 Codex improved human preference scores on mobile websites. In addition, when GPT-5 Codex is used ...

The ChatGPT maker claimed a SWE-bench Verified benchmark success rate of 74.5%, with refactoring performance improving to ...

In addition to measuring overall user and usage growth, OpenAI's paper also breaks down total usage based on when its ...

Unlike GPT-5, which is built as a general-purpose AI model, GPT-5-Codex is optimized for what OpenAI calls “agentic coding," ...

OpenAI has launched GPT-5 Codex, an AI model built for software development. It can write, debug, and review code, working ...

New reports added fresh data to the worldwide debate over the AI's economic impact and testing the companies’ own predictions ...

It is the first study to use internal ChatGPT message data, for which OpenAI said privacy-preserving techniques were used to ...

OpenAI has upgraded its AI coding agent, Codex, to its latest GPT-5 AI model, CEO Sam Altman announced on X. Called GPT-5 ...