Adversarial algorithms can systematically probe large language models like OpenAI’s GPT-4 for weaknesses that can make them misbehave.

IBM and Meta launched the AI Alliance today along with over 50 global organizations to "support open innovation and open...

Researchers at Long Island University found that ChatGPT only correctly answered 10 out of 39 medical-related questions in a study.

Microsoft Corp. is updating Microsoft Copilot, its collection of conversational search services, with new features that will enable users to...

Tech giants form yet another coalition to champion open source AI innovation and development.

ByteDance is planning to build a platform that will let users create their own chatbots, per a report from the...

Together, the companies will develop develop a new generative AI video model: Runway Getty Images Model (RGM), targeting Hollywood &...

Jason Koebler / 404 Media: Asking ChatGPT to repeat words “forever” now violates OpenAI's terms of service, after Google researchers...

Cooley's Bobby Ghajar, Mark Weinstein and Judd Lauter secured a ruling from a judge in San Francisco dismissing a significant...

Donald Trump attacked the Lincoln Project on Monday night, claiming the political advocacy group is using artificial intelligence to make...