News & analysis, rated.
Every story from the daily brief, each rated so you know what matters. Hand-picked by the desk, not an automated firehose.
Browse the news archive →Google's AMIE matches primary-care doctors in live video consults
Google Research published a randomized study in which AMIE (Video) matched or exceeded 10 board-certified primary care physicians across diagnosis, reasoning, and communication in simulated real-time video consultations.
Researchers warn Medicare's AI-device pay code will drive overuse
STAT reported Wednesday that Medicare's NTAP payments — designed to sweeten hospital adoption of newly authorized AI devices — could incentivize overuse, according to researchers reviewing the program.
OpenAI ships a frontier model built for exploit development, gated by partner tier
OpenAI restructured its Daybreak cybersecurity program into two access tiers on Monday and released GPT-5.6-Cyber, a model purpose-trained for offensive security work.
IBM commits $240M to a Together AI inference cluster on Blackwell
IBM and Together AI signed a multi-year $240M agreement Tuesday to build a dedicated inference cluster on IBM Cloud using Nvidia HGX B300 systems.
Anthropic will watermark all new Claude output worldwide for EU compliance
Anthropic committed to embedding invisible text watermarks and C2PA metadata into files generated by Claude models launched from August 2, applied globally rather than only inside the EU.
OpenAI's CFO publishes a five-lesson playbook for AI-native finance
OpenAI CFO Sarah Friar published a Friday post laying out five lessons from rebuilding finance around AI, with a stated goal of zero-day close and continuously updated forecasting.
FT: India's $250B IT engine faces its first real coding-agent stress test
The FT reported maturing coding agents are compressing the labor demand underpinning India's IT services industry — the clearest large-scale AI substitution signal yet in a named geography.
Meta open-sources Muse Glimmer and pledges Muse Spark 1.2 weights
Meta shipped a 30B open-weight agentic model that runs on a laptop and committed to release the weights of its flagship — the first US frontier lab to open its top-tier model in a year.
Anthropic, Macquarie and GIC form "Theseus" data-center vehicle
Anthropic announced a joint venture with Macquarie Asset Management and Singapore's GIC to build and lease US data centers, with Anthropic as anchor tenant.
Intel sells $15B in stock — first share offering since 1971 IPO
Intel announced a $15B underwritten public offering — Bloomberg says its first common-stock sale since going public in 1971 — citing AI compute demand and physical-AI silicon.
OpenAI cannot rule out Astra crosses "Critical" cyber threshold
OpenAI disclosed Friday that its upcoming Astra model may meet the Critical cybersecurity level in its Preparedness Framework, paused some internal work, and tightened access — CNBC reported Monday.
EU AI Act transparency rules go live, but high-risk provisions slip to 2027
Article 50 transparency obligations became enforceable across the EU on August 2, while the AI Omnibus quietly pushed the Act's high-risk system deadlines out to December 2027.