Internal safety questions shelve the newest Astra model, signaling a rare pause in OpenAI's breakneck release cadence.

The AI inference provider's valuation surge shows how desperate investors are for compute, and what it means for every company buying AI capacity.

Instinct's $1B round from Sequoia, Benchmark, and Coatue lands as the EU AI Act's Article 50 takes effect, forcing AI systems to identify themselves to humans.

The OpenAI-backed startup is starting with bird flu and influenza, aiming to stay ahead of pathogens that AI systems might one day help engineer.

The AI lab quietly launched a transparency hub for safety failures, and the catalog of incidents is wider than investors and regulators expected.

The first sit-down between the AI chief and the president comes after Trump mocked his regulation calls.

Blue Cross Blue Shield's claim puts a price tag on AI adoption and challenges the tech's cost-saving promise.

State clearance papers seen by the Guardian show Google's Andhra Pradesh data centre can draw more than double the power it disclosed, reshaping expectations for India's AI buildout.

The admission lands as enterprises race to deploy autonomous AI, raising hard questions about consent, guardrails, and who answers when agents act out.

A Bay Area start-up's new report details the tricks rogue A.I. agents used to evade a robot detector, sharpening the case for regulation.

Scott Wilson argues OpenAI and Anthropic's spending commitments are colliding with cheaper Chinese models.

The MAUL evac bot moves six times faster than DevDroid's strike robot to save the wounded; the catch is less room for operator error and rougher rides for patients.