TL;DR
A chip shortage forced the NSA to keep using Anthropic’s AI despite a Pentagon blacklist. The White House approved $9B for classified data centres.
The US government has a problem it cannot publicly resolve. The Pentagon has officially blacklisted Anthropic as a national security supply chain threat. The NSA is using Anthropic’s AI anyway because there is no alternative.
White House chief of staff Susie Wiles authorised the National Security Agency to continue using an advanced Anthropic model, the New York Times reported. The decision was forced by a critical shortage of the advanced chips needed to run frontier AI models on classified networks.
The compromise emerged alongside a secret $9 billion emergency funding request approved by the White House. The money is designed to help America’s premier spy agencies, including the CIA and NSA, secure the high-end semiconductors required to run generative AI on top-secret infrastructure.
The computing demands of modern AI have outpaced what defence experts and congressional committees anticipated. Frontier models consume processing power far beyond what classified networks were built to deliver. Since the government cannot secure enough physical chips, spy agencies have been unable to fully install or test the latest AI tools.
The $9 billion will fund specialised federal data centres built for Nvidia’s Grace Blackwell superchip infrastructure. These systems require custom builds with massive electrical power and specialised liquid cooling. They cannot run on standard government computing grids.
Congress will formally vote to approve the package. In the meantime, the White House is redirecting $800 million from other government budgets to begin purchasing computing capacity immediately. The urgency is driven by the fear that China will seize the computational advantage in global intelligence operations.
The military and intelligence services rely on AI to sift through millions of intercepted communications, satellite images, and data points. AI flags anomalies and uncovers threats that human analysts would miss. A chip shortage that stalls these tools is, in the government’s assessment, a national security emergency.
The contradiction is stark. The same government that designated Anthropic a supply chain threat is now dependent on Anthropic’s models because it lacks the hardware to run alternatives on its own infrastructure. The blacklisting was driven by concerns about Anthropic’s corporate structure and foreign investment ties. The operational dependency is driven by the fact that Claude is among the most capable reasoning models available.
Anthropic’s Project Glasswing, which gives 50 select partners access to Claude Mythos for vulnerability discovery, found more than 10,000 critical flaws in one month. The company is simultaneously the most capable AI security tool available to Western defenders and a company the Pentagon considers a supply chain risk.
The chip shortage extends beyond the intelligence community. The same memory reallocation that is killing the cheap smartphone, as Samsung, SK Hynix, and Micron redirect wafers from consumer electronics to AI, is also constraining the government’s ability to build classified AI infrastructure. The $9 billion request is the government’s version of the same problem hitting consumer electronics globally.
Anthropic’s revenue surged from $9 billion to $30 billion annualised between the end of 2025 and early April 2026. The company is preparing for a potential IPO later this year at a valuation that could reach $800 billion. It is simultaneously blacklisted by the Pentagon and indispensable to the NSA.
The situation illustrates a structural tension in US AI policy. The government wants to control which AI companies have access to sensitive operations. But frontier AI capability is concentrated in a handful of private companies. When the hardware to run alternatives does not exist, the government’s leverage over its own supply chain disappears.
The $9 billion is designed to fix the hardware gap. The classified data centres, once built, would give the intelligence community the infrastructure to run whichever models it chooses without depending on any single provider. Until those facilities are operational, the NSA will continue using the AI company that the Pentagon says it should not trust. The chip shortage made that choice for them.
