ChatGPT should feel “smarter and more accurate” starting today, according to OpenAI. That’s because the company is replacing ...

Learn how the new OpenAI Codex acts as an AI chief of staff to automate workflows, manage emails, and navigate your computer ...

ChatGPT's default model has been updated to GPT-5.5 Instant, a model that brings accuracy improvements with fewer ...

OpenAI’s newest default model for ChatGPT might not make stuff up as much. Hallucinations have been an ongoing problem for AI ...

OpenAI has traced ChatGPT’s bizarre goblin fixation to training gone awry, but the creatures keep escaping—as Adam Engst discovered when goblins popped up in a conversation about a conference ...

OpenAI details all the improvements users can expect to experience when interacting with ChatGPT's default model.

GPT-5.5 Instant is now the default ChatGPT model. It makes fewer things up, answers more concisely, and finally remembers ...

The company said the model reduces hallucination in sensitive areas such as law, medicine, and finance, while maintaining the ...

ChatGPT 5.5 has finally "dropped the clipboard," says one user who's been missing ChatGPT 4o's engaging personality. Is ...

But the most realistic deepfakes I was able to create did not involve politicians or celebrities. They mostly did not depict ...

Following its recently announced partnership with Novo Nordisk, OpenAI is introducing a new reasoning model, GPT-Rosalind, to support research in biology, drug discovery and translational medicine.

Named after the crystallographer who helped reveal the structure of DNA, GPT-Rosalind is OpenAI’s first domain-specific model series, fine-tuned for biochemistry, genomics, and protein engineering.