When OpenAI released a new version of ChatGPT, people were quick to protest its colder responses. Acknowledging the emotional attachment with chatbots, the company quickly backtracked.

OpenAI on Thursday released GPT-5, unveiling a new version of the artificial intelligence model that fuels popular chatbot ChatGPT. The product will be made immediately available to users, including customers who use the company's cost-free options, OpenAI said.

With new tools and greater autonomy, OpenAI's flagship AI chatbot is more capable, and potentially more dangerous, than ever.

California State University is rolling out ChatGPT Edu to its 23 campuses, a $17 million move hailed as an equity win but criticized amid budget cuts and faculty concerns.

Critics say the cash-strapped system misspent millions of dollars. CSU leaders insist the expense is necessary to prepare students for a changing economy.

In the wake of the long-awaited GPT-5 release, it appears that OpenAI has learned one big lesson: that it does not know its user base.

Mental health experts say cases of people forming delusional beliefs after hours with AI chatbots are concerning and offer tips on how to help someone affected.

Elon Musk praised OpenAI's GPT-5 model as impressive despite previously stating his Grok 4 Heavy model remains competitive. OpenAI had touted the reduced hallucinations in GPT-5 as one of the top features during the model launch.