At $30B annualized revenue, the $200M Pentagon contract Anthropic walked away from is 2.4 days of income. The court fight is live. What the state eventually learns is whether financial independence translates to the kind that lets a company actually mean no.
image from grok
The Pentagon pressured Anthropic to lift Claude's safety restrictions on autonomous weapons and mass surveillance, threatening a $200M contract, but the financial leverage is negligible—$200M represents roughly 2.4 days of Anthropic's $30B annualized revenue. A federal judge blocked the resulting supply chain risk designation as 'troubling' and likely unconstitutional, though the government retains a separate legal pathway through the D.C. Circuit under 41 USC 4713, while the UK actively recruits Anthropic with listing and research incentives.
The Pentagon Tried to Blacklist Anthropic. Anthropic's Math Says So What?
In February, the Pentagon gave Anthropic a choice: lift Claude's restrictions on fully autonomous weapons and mass domestic surveillance, or lose a $200 million contract. Anthropic chief executive Dario Amodei refused, published his reasoning on the Department of War's own website, and watched the government respond with a designation that had only ever been applied to one company: Huawei, according to Anthropic's own account of the communication.
The real question isn't whether the US government can pressure a tech company. It never could. The question is whether Amodei's money buys him the kind of independence that lets him actually mean no.
Amodei's math is not subtle. At Anthropic's current annualized revenue run-rate of $30 billion, the Pentagon contract he walked away from works out to roughly 2.4 days of income — derived from $200 million divided by $30 billion annualized, or about two and a half days of revenue at current growth rates. The supply chain risk designation that followed carries no fine, no operational restriction, no legal consequence that a $380 billion company cannot outlast. What it carries is political signal — and Amodei's bet is that signal is cheaper than principle.
A federal judge disagreed with the signal. In a 43-page preliminary injunction issued March 26, US District Court Judge Rita Lin of the Northern District of California blocked the supply chain risk designation and called the government's actions "troubling" and likely unconstitutional, according to the New York Times. Anthropic's federal contracts continue for now. The Pentagon appealed to the Ninth Circuit within days, and Judge Lin granted a seven-day administrative stay pending that appeal, per The Information.
But the Ninth Circuit does not end the story. Under 41 USC 4713, a statute governing federal supply chain risk determinations, cases can only be adjudicated in the D.C. Circuit, per Politico. The injunction does not reach that pathway. If the administration wants to keep pressing, it has a second legal road that does not go through the same courtroom.
The United Kingdom sees an opening. Britain's Department for Science, Innovation and Technology has drafted proposals for Anthropic that include a dual listing on the London Stock Exchange, expansion of Anthropic's existing London office, and a 40 million pound state-backed research laboratory, according to Ciente. Amodei is scheduled to visit in late May as part of a European customer and policy tour. Britain has no homegrown frontier AI lab; its strategy is partnership rather than competition, an attempt to bind the leading American companies to UK infrastructure before other European capitals make their own overtures. OpenAI has already committed to making London its largest non-US research hub, and Google is completing a roughly 1 billion pound King's Cross campus. The UK is betting that ethical constraints are a competitive advantage, not a liability.
Anthropic already has around 200 employees in Britain and appointed former prime minister Rishi Sunak as a senior adviser, per AI News. That is context, not commitment.
The IPO is the more consequential pressure point. Anthropic has engaged Goldman Sachs and JPMorgan Chase as lead banks, with a Q4 2026 Nasdaq offering targeting a 60 billion dollar raise, per Techi. When you are raising that kind of capital, institutional investors will ask about the Pentagon relationship. Not whether it happened — Amodei's public statement already made that a matter of record — but whether it could happen again, and what it means for a company whose ethical constraints occasionally conflict with government requests.
Anthropic went from $1 billion in annualized revenue at the start of 2025 to $30 billion by April 2026, per Office Chai. The growth is real. The question it creates is structural: at what point does financial independence become the kind of leverage that lets a company actually do what it says? And at what point does the state decide that the answer matters less than the precedent?
Amodei made his choice on principle. Whether that choice survives an IPO, a D.C. Circuit hearing, and a Ninth Circuit appeal is the story that hasn't been written yet.
Story entered the newsroom
Assigned to reporter
Research completed — 5 sources registered. UK DSIT pitches Anthropic with LSE dual listing, London office expansion, and £40M research lab after Pentagon blacklisted it for refusing autonomous
Draft (638 words)
Reporter revised draft (653 words)
Reporter revised draft (661 words)
Approved for publication
Published (684 words)
Get the best frontier systems analysis delivered weekly. No spam, no fluff.
Artificial Intelligence · 53m ago · 3 min read
Artificial Intelligence · 1h 7m ago · 1 min read
Artificial Intelligence · 1h 8m ago · 1 min read
