White House is Divided Over AI Policy | $9B Approved for Spy Agencies to Have Best Chips

The White House just approved $9 billion for the CIA and NSA to buy Nvidia's most advanced AI chips and run them at scale. That's the headline. But it gets even more interesting.

Back in July 2025, Anthropic signed a $200 million contract with the Pentagon — with two hard limits: no mass surveillance of Americans, no autonomous weapons. Six months later, the Pentagon asked them to drop those limits. Anthropic said no.

So the Pentagon labeled them a "supply chain risk to national security" — a label previously reserved for foreign adversaries like Huawei. First time it was ever used on an American company.

And here's the paradox: while Anthropic sat on that blacklist, the NSA was actively running Mythos! — Anthropic's most locked-down cybersecurity model, available to roughly 40 organizations on the planet.

In this video I break down the full timeline, the court battles, and why the new NSA contract is being written on Anthropic's terms — not the Pentagon's.

This is about more than chips. It's about who decides how AI gets used inside a government.

⬇️ What do you think — did Anthropic hold the line on principle, or on leverage? Please let me know. 🙏🏻
—
🔔 Subscribe for weekly breakdowns on the AI news that actually matters for regular people — not the hype, not the noise.
—
#AINews #Anthropic #Pentagon #AIPolicy #NationalSecurity #CIA #NSA #AISafety #NvidiaBlackwell #AIChips #TechNews #AIGovernment #MariaBlairTech #MariaTalksTech

Видео White House is Divided Over AI Policy | $9B Approved for Spy Agencies to Have Best Chips канала Maria Talks Tech