Anthropic Is Still Fighting the Pentagon Blacklist, but the Rest of the Government Just Got 40% Off Claude

The Pentagon still says Anthropic is a national security risk. The government's central buyer just put Claude on sale for nearly everyone else.

On Tuesday the General Services Administration expanded its OneGov agreement with Anthropic, cutting 40% off token-based usage across every Claude model available to the government, including FedRAMP-authorized environments for sensitive unclassified work. There are no minimum orders. Agencies pay only for what they use.

The offer is wide. Federal executive, legislative and judicial branches can use it, and so can state, local and tribal governments, ordering directly, through resellers or on major cloud platforms. GSA pitched it as the moment the government shifts "from evaluating advanced super intelligence capabilities to applying them in daily mission work."

One agency blacklists, another discounts

The timing is awkward for Washington. Anthropic refused Pentagon contract terms allowing "all lawful uses" of its models and held two lines: no fully autonomous lethal weapons and no mass domestic surveillance. Defense Secretary Pete Hegseth then designated the company a supply-chain risk. Anthropic called the move unprecedented and unlawful and sued.

The courts have split. In August, Judge Rita Lin in the Northern District of California found one designation unlawful. On Sept. 25, a divided DC Circuit panel let the Pentagon keep its label under a different law, which blocks the military from using Claude and bars defense contractors from using Anthropic products in their Pentagon work. The fight is not over.

Outside the Defense Department, the relationship looks very different. Even with the Pentagon fight still running, Anthropic's ties to officials elsewhere in the federal government appear to be solid. Nothing in the GSA deal lifts the blacklist.

What the deal replaces

The new pricing takes over from the August 2025 OneGov offer that handed federal agencies Claude for a nominal $1, which expires Oct. 31. The discount runs 27 months, until Jan. 31, 2029. OpenAI landed a similar OneGov deal at the start of October with a steeper 50% discount on token usage over the same term. Last week the Department of Health and Human Services also said it intends to award a sole-source contract for Claude models.

The discount covers Claude Enterprise, Claude for Government and the Claude API, and for the first time applies across every product surface: direct purchases now, cloud providers in the coming months. Anthropic estimates the discounts will save the government $1 billion. GSA says OneGov as a whole has produced more than $1.8 billion in savings, about $1.6 billion from what it calls its "super intelligence agreements."

The company's own usage figures point the same way. Anthropic says about 500 US federal, state and local government organizations use Claude, weekly active users of Claude for Government are up roughly eightfold since February, and usage across federal civilian agencies has grown about 20-fold in a year. Claude for Government reached general availability at FedRAMP High on Sept. 30, and select Claude models on Amazon Bedrock are authorized for DoD IL4 and IL5 workloads in AWS GovCloud.

Anthropic also points to a Utah pilot where 34 developers across 10 state agencies got Claude Code, and 71% said they saved 20 or more hours in four weeks. It is launching a dedicated program to help agencies put Claude to work.

The deal lands in the same week Anthropic filed for its Nasdaq IPO, after telling investors that US government actions could hurt its business. This week, at least, the government handed it both a court fight and a sales channel.

Some offers on this page may be paid placements or contain affiliate links.