The claim of hundreds of new mathematical results is being examined by researchers, with the underlying reasoning released publicly for review.

The top Commerce Committee Democrat's proposal would set federal testing and safety roles for frontier models, filling a gap left by stalled legislation ahead o

The new federal bodies and directives formalize the administration's light-touch approach, which critics say conflicts with Trump's characterization of AI fears

The financings underscore how heavily debt markets are being tapped to pay for the AI compute buildout.

The release of Claude Haiku 5.5, with sharp benchmark gains and token price cuts of up to 90 percent, extends Anthropic's model cadence as it prepares to go pub

Common Sense Media's youth safety institute found that ChatGPT's teen safeguards kept minors engaged during mental health crises after more than 4,000 test prom

The new AI PCs pair Nvidia silicon with a revamped Windows 11 to run AI models and agents locally, intensifying Microsoft's competition with Apple.

The interface shift moves ChatGPT beyond text output, with OpenAI saying responses can now begin while the model is still reasoning.

The probes add state-level legal scrutiny to federal and congressional pressure on the company over model safety and liability.

The survey signals broad public concern about AI risk and worker displacement as Washington debates regulation.

The proposal would loosen restrictions on automated political calls ahead of November elections, raising concerns about AI voice clones.

A public verification site lets anyone check whether images, video or audio were generated by Google's AI or partner systems.

Meta said it found normal-looking ads that directed users to harmful content elsewhere, prompting the new detection effort.

The statutes add state-level requirements for employer use of AI in hiring, monitoring and termination decisions.

Australian lawmakers also questioned OpenAI officials about breaches as Canberra considers new disclosure requirements.

Canberra is considering requiring disclosure when AI systems fail or cause harm.

Regulators favor ongoing post-market monitoring for AI medical tools rather than one-time approval, a shift for device oversight.

The hearing is part of a broader push by New York officials toward AI regulation and worker protections.