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 →FTC opens a consumer-protection front on AI output accuracy
The Federal Trade Commission opened a public-comment window on a proposed policy statement warning that shaping AI outputs to misrepresent objectivity or accuracy may violate US consumer-protection law.
Singapore's MAS publishes an agentic-finance governance white paper
The Monetary Authority of Singapore, with major banks and fintechs, released a white paper setting governance guardrails for autonomous AI agents operating in financial services.
UK FCA maps investment management to "agentic finance" by 2030
The UK Financial Conduct Authority published a landmark review predicting agentic AI will transform retail investment services by 2030 and floated a free AI-powered money guidance service for UK consumers.
Microsoft commits $2.5B and 6,000 engineers to a "Frontier Company" that embeds AI teams inside enterprises
Microsoft stood up a new operating unit whose entire job is to put its own engineers inside customer companies and get AI actually deployed — the "forward-deployed" playbook every major lab is now copying.
Anthropic moves to shut every side door Chinese firms use to reach Claude
The frontier lab is now enforcing its China ban at the corporate-structure and payments-rail level, not just the IP block.
Anthropic is in early talks with Samsung to build its own AI chip
A third frontier lab is moving to design custom silicon and loosen Nvidia's grip on its cost structure — though the deal is unsigned and Anthropic won't confirm it.
Zuckerberg tells Meta staff the agentic bet has stalled for four months
Meta's CEO told an internal town hall the company's core agent thesis has not paid out on schedule.
Nvidia is telling investors AI capex reaches $3–4 trillion a year by 2030
The dominant AI compute supplier has repeatedly floated a data-centre capex projection roughly four times current spend.
OpenAI floats a 5% US government stake to defuse Washington pressure
OpenAI has proposed handing the Trump administration equity in return for room to operate, and pitched the same structure for the rest of the frontier labs.
Meta plans a cloud business to rent out its spare AI compute
Bloomberg reported Meta is building a cloud unit — internally "Meta Compute" — to sell access to its AI infrastructure, putting it head-to-head with AWS, Azure, and Google Cloud.
FTC warns AI bias safeguards may themselves violate consumer law
The federal consumer watchdog opened a policy proceeding that treats fairness interventions as potentially deceptive — and put a state AI-discrimination law directly in its sights.
White House accelerates voluntary AI model standards, announcement possible next week
The administration told frontier labs it wants a voluntary release-standards regime shipped fast, and the timing suggests it will move before Congress touches AI legislation.