Anthropic drops Claude Gov for natsec customers, hastening the public sector AI race
Anthropic announced Thursday that it is releasing Claude Gov to U.S. national security customers, an exclusive set of artificial intelligence models that is already in the hands of some government agencies.
The new government AI product, which Anthropic detailed in a blog post, comes as several companies compete to sell emerging technology tools to federal agencies. At stake is new business — and the prestige of working on serious government missions.
Earlier this year, OpenAI released ChatGPT Gov, a specialized version of its chatbot. The company recently announced that the Space Force and the Air Force Research Lab were the product’s first customers. Meanwhile, Department of Government Efficiency representatives have reportedly been advocating for the use of Grok — the chatbot produced by Elon Musk’s firm xAI — within the Department of Homeland Security.
AI companies have been partnering with cloud providers already approved through the FedRAMP-process, which is now undergoing a major overhaul, in order to boost their technology’s presence in government.
Anthropic said the models available under Claude Gov are designed to handle classified materials, perform better in languages and dialects “critical to national security,” and have a greater understanding of the intelligence and defense contexts.
“What makes Claude Gov models special is that they were custom-built for our national security customers,” Thiyagu Ramasamy, the head of public sector work at Anthropic, said in a statement. “By understanding their operational needs and incorporating real-world feedback, we’ve created a set of safe, reliable, and capable models that can excel within the unique constraints and requirements of classified environments.”
Last November, Palantir and Anthropic announced a partnership to share Claude systems with the U.S. government through Amazon Web Services. Claude was also one of the tools approved for conditional use at the Department of Homeland Security, before the agency revoked access to those tools and directed staff to internal tools, including DHSChat.
