Anthropic collaborates with AWS and Palantir to offer AI to Defense Clients

Anthropic announced that it is collaborating with Amazon Web Services (AWS) and data analytics company Palantir to make its Claude family of AI models available to U.S. intelligence and defense agencies.

The announcement comes as more and more AI suppliers seek to sign contracts with US defense clients for financial and strategic reasons. While OpenAI is looking to forge a closer partnership with the U.S. Defense Department, Meta recently announced that it is making its Llama models available to defense partners.

According to Kate Earle Jensen, head of sales for the company, Anthropic’s partnership with Palantir and AWS would “operationalize the use of Claude” within Palantir’s platform by utilizing AWS hosting. Claude is currently usable in Palantir Impact Level 6 (IL6), the defense-accredited environment, and was made accessible on Palantir’s platform earlier this month.

Also Read: Anthropic Introduces Claude 3.5 Sonnet, Haiku & AI Computer Use to Boost Efficiency

Systems holding data that is considered vital to national security and needing “maximum protection” against manipulation and unauthorized access are designated for the Defense Department’s IL6. In IL6 systems, information can be as high as “secret,” which is one level below top secret.

“We take great pride in leading the way in introducing responsible AI solutions to U.S. classified environments, improving operational efficiencies and analytical capabilities in critical government operations,” Jensen stated. “U.S. defense and intelligence agencies will have access to Claude within Palantir on AWS, which will give them strong AI tools that can quickly process and analyze enormous volumes of complex data.” This will significantly enhance intelligence analysis, facilitate decision-making for officials, simplify resource-intensive duties, and increase departmental operating efficiency.

To increase the number of clients it serves in the public sector, Anthropic introduced a few Claude models to AWS’ GovCloud this summer. AWS’ GovCloud service was created specifically for cloud workloads used by the US government.

Also Read: Anthropic Claude 3.5 Sonnet and Haiku: Benchmark, Capabilities and Key Features

Compared to OpenAI, Anthropic has positioned itself as a vendor that is more concerned about safety. However, according to the company’s terms of service, its products can be used for things like “identifying covert influence or sabotage campaigns,” “providing warning in advance of potential military activities,” and “legally authorized foreign intelligence analysis.”

According to Anthropic’s terms, “[We will] tailor use restrictions to the mission and legal authorities of a government entity” depending on things like “the extent of the agency’s willingness to engage in ongoing dialogue.” It points out that the terms do not apply to AI systems that it believes “substantially increase the risk of catastrophic misuse,” exhibit “low-level autonomous capabilities,” or have the potential to be used for malicious cyber operations, censorship, disinformation campaigns, weapon design or deployment, or domestic surveillance.

AI undoubtedly piques the curiosity of government organizations. The Brookings Institute reported a 1,200% increase in government contracts connected to AI in March 2024. However, some sectors, including the US military, have been sluggish to embrace the technology and continue to doubt its return on investment.

Also Read: Nvidia’s New Llama-3.1-Nemotron AI Model Takes on OpenAI & Anthropic with Superior Performance

According to reports, Anthropic, which recently opened offices in Europe, is negotiating a fresh round of funding at a potential $40 billion valuation. The total amount raised by the corporation, including forward pledges, is around $7.6 billion. By far its biggest investor is Amazon.