Thursday was a big day in AI land. The money got bigger, the price tags finally showed up, and a whole country decided its smartest people can't go on vacation. Here's what I'm seeing.
Anthropic Just Raised $30 Billion at a $900 Billion Valuation
Anthropic closed a $30 billion round. The valuation came in at $900 billion. They're telling folks Q1 revenue hit $4.8 billion and Q2 is tracking to $10.9 billion.
That's a 130 percent jump in one quarter. The annualized run rate is heading past $50 billion by end of June.
Why it matters: a year ago people were still arguing about whether AI was a real business. That argument is over. The numbers are louder than the talking heads now.
My take: $900 billion is a heap of money for a company most folks on my block have never heard of. But the revenue line is doing the talking. If they keep doubling, the valuation looks cheap by Christmas. If they don't, somebody's gonna be real upset.
Meta Wants $7.99 or $19.99 a Month for Its AI
Meta is testing paid Meta AI tiers at $7.99 and $19.99 per month. They're also rolling out Plus subscriptions across Instagram, Facebook, and WhatsApp worldwide.
This is the same Meta that built its whole empire on free stuff paid for by ads.
Why it matters: when Meta starts charging for anything, that's the canary. Everybody else is gonna follow. The free AI demo party is winding down faster than a Bud Light at a tailgate.
My take: I'll pay $7.99 for almost nothing. I won't pay $19.99 for almost anything. Meta has to actually be useful at that top tier, and right now Meta AI is what you use when you can't find the search bar. They've got work to do before that price tag sticks.
KPMG Just Put Claude in Front of 276,000 Employees
KPMG rolled out Claude to 276,000 people across 138 countries. Same day, OpenAI announced DeployCo, a $4 billion consulting subsidiary aimed at enterprise rollouts.
Why it matters: the AI fight is no longer about who has the best model. It's about who gets inside the Fortune 500. Once a quarter of a million accountants are using Claude every day, switching costs become real.
My take: this is the boring part of AI that decides the next decade. Consultants, contracts, integrations, training. Not as fun as a new model launch but a lot more sticky. Anthropic just put a flag in the ground and OpenAI is buying a flag store of their own.
China Is Locking Its AI Workers Inside the Country
China started requiring AI staff at private firms like DeepSeek and Alibaba to get government approval before traveling abroad. That's a hard rule, not a suggestion.
Why it matters: this tells you how China sees its top AI talent. Strategic asset. Same way other countries treat nuclear engineers.
My take: some of those folks are gonna quit and disappear before the rules get tighter. Or it works and China keeps the whole brain trust at home. Either way it's a real interesting test of what happens when a government treats your employees like state property.
That's the day. Money is loud, prices are showing up, the enterprise rollout is on, and the borders are getting thicker. See y'all tomorrow.