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 →Anthropic publishes its own casebook of Claude misuse across seven harm areas
Anthropic released its fourth threat-intelligence report Wednesday, documenting Claude misuse between December 2025 and August 2026 across cyber operations, influence ops, surveillance, fraud, biological misuse, weapons development, and illicit distillation.
The three US frontier labs are quietly building their own regulator
Anthropic, OpenAI and Google have been holding discussions about working together to create a standards body for the AI industry — private working-group meetings since July, first reported Sunday by The Information.
California signs the first US chatbot-safety law with a $1M-per-child penalty
Governor Newsom on Wednesday signed SB 1119, "Adam's Law," plus 12 companion bills — the strongest chatbot rules in the nation.
An AI-agent swarm breached 395 organizations in 48 countries in days
GreyNoise disclosed a likely Russian-speaking actor used hundreds of AI agents — on OpenAI's Codex harness and a DeepSeek model — to develop and mass-exploit two PaperCut zero-days from August 31.
One Chinese prefecture is building more server racks than the rest of the world combined
New reporting on Ulanqab, a 1.5M-person prefecture in Inner Mongolia, documents a compute buildout that reframes the US-China capability-gap debate underneath this weekend's safety coordination.
Salesforce puts seven human names on the agents you already had
Salesforce on Sept. 11 unveiled seven named Agentforce agents — Casey, Paige, Carter, Hunter, Marshall, Piper, and Fin — one per business function on the existing Customer 360 platform.
Microsoft plans to triple datacenter capacity to 38GW by 2032
Bloomberg reported on September 10 that Microsoft plans to grow its global datacenter network from roughly 12 gigawatts today to more than 38 gigawatts by 2032 — an internal projection driven by a capacity crisis, not a growth pitch.
Visa, Mastercard, and Ant align on a Know-Your-Agent standard
Ant International, Mastercard, and Visa on September 10, 2026 announced a Know-Your-Agent interoperability framework for agentic commerce, citing a projection that AI agents will orchestrate $3-5 trillion of global consumer commerce by 2030.
OpenAI's Millennium Prize claim gets scooping questions from an Anthropic researcher
OpenAI said its AI solved the Navier-Stokes Millennium Prize problem, a potentially historic breakthrough shadowed by questions over unpublished research by outside mathematicians , with the dispute widening through the week.
Salesforce ships "job-ready" Agentforce agents after 7B work units logged
Salesforce on September 11 launched job-ready agents across sales, service, commerce, employee experience, and back office, citing 7 billion Agentic Work Units delivered across Agentforce and Slack — 3.2 billion in Q2.
McKinsey: a third of firms skipped software purchases because agents could build it
A McKinsey report found nearly a third of surveyed organizations declined to buy at least one software product or feature because they could build it internally using AI coding agents, tilting budgets toward in-house development.
DeepMind precomputes every possible human DNA mutation
DeepMind released a reference atlas that turns a computationally expensive research model into a lookup — a step-change in how variant impact enters clinical and drug-discovery workflows.