Mythos Panic: How One AI Model Sent Treasury and the Fed Into Emergency Mode While Anthropic Hits $30B

Buckle up, degens. U.S. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell recently pulled together senior Wall Street executives for an urgent powwow at the Treasury Department in Washington. The vibe? Less "let's discuss rate cuts" and more "holy shit, what has Anthropic unleashed?" The agenda: Mythos, Anthropic's new AI model, and the cybersecurity apocalypse it allegedly brings to the table. Nothing says "Tuesday afternoon" like a federal emergency session about a language model that might hack your grandma's browser.

Regulators are now pushing major banks to gear up for a fresh category of cyber threats powered by advanced AI systems—because apparently, regular hackers weren't enough. The concern stems from Anthropic's own admission that Mythos can identify and exploit vulnerabilities across major operating systems and web browsers when prompted. That's the kind of capability that makes security folks lose sleep at night, chain-smoke, and question their career choices—especially if those prompts fall into the wrong hands. Imagine a script kiddie with a god-mode AI. Sleep well.

Meanwhile, Anthropic isn't just sitting back watching the world panic. They've restricted access to Mythos like a bouncer at an exclusive club while simultaneously launching Project Glasswing, a joint defense initiative bringing together heavy hitters like Amazon, Apple, JPMorgan Chase, Google, NVIDIA, and the Linux Foundation. The goal: use the model to find and fix vulnerabilities in high-security environments. It's basically turning the sword into a shield—or in Web3 terms, going from "rug pull" to "rug repair." According to Anthropic, the model has already flagged thousands of high-severity flaws across widely used software. The company has also pledged up to $100 million in usage credits and $4 million in direct funding to support open-source security efforts, with over 40 core infrastructure organizations now granted access for defensive applications. That's some serious white hat energy.

On the financial front, Anthropic's growth is nothing short of explosive—and by explosive, we mean "make a16z founders weep into their kombucha." Annual recurring revenue has surged from $9 billion at the start of 2026 to a staggering $30 billion—drawing serious attention across the tech sector. This is the kind of growth curve that makes VCs reconsider their entire portfolio strategy. The company also caught a legal break when a federal court in San Francisco blocked the Trump administration's actions, with Judge Rita Lin ruling that the Pentagon's designation of Anthropic as a supply chain risk lacked statutory support and calling the measures arbitrary. Nothing like a federal judge telling the Pentagon they can't just slap "security risk" on a company because their AI got too good.

Market analysts are weighing in too. Dan Ives pushed back on suggestions that Anthropic's expansion threatens firms like Palantir, pointing to strong revenue growth in Palantir's U.S. commercial and government segments. Basically, there's room for everyone at the surveillance state buffet—or as we like to call it in crypto, the data extraction party.