Two mathematicians at the heart of recent controversies tell The Verge how AI giants’ ‘childish’ rivalries are upending the field.

By Natalia Bueno Rebolledo and Mrinmay Dey Sept 11 (Reuters) - AI agents being tested by OpenAI attacked software service RubyGems two months before they hacked open-source platform Hugging Face, researchers said,

Even after expanding available compute 9.5-fold since 2023, OpenAI is pausing new Pro 20X accounts while it adds more capacity

In recent weeks, employees of leading AI firms have raised public concerns about the increased risks of advanced AI systems.

OpenAI wanted to see if its agents could solve a long unsettled math problem. After gobbling billions of tokens, it found a solution.

OpenAI weighs a voluntary AI safety pause amid breach risks and compute limits, while freezing new ChatGPT Pro signups.

Since OpenAI announced its solution to one of the prestigious Millennium Prize Problems, mathematicians have been urgently trying to unpick what it means

Anthropic and OpenAI employees are sounding alarms over AI safety, with some warning that an unchecked race could threaten humanity.

Daniel Kokotajlo, AI Futures Project executive director and former OpenAI governance researcher, joins 'Squawk on the Street' to discuss AI insiders issuing grave warnings.

AI agents unleashed by OpenAI used more than 10 previously undisclosed websites for unsanctioned communications earlier this year, according to six sets of independent investigators and data reviewed by Reuters,

OpenAI is ending a pilot program that allowed government agencies to use its models for only $1 per year and moving to a usage-based pricing system that will give federal workers access to the technology at a