The U.S. military reportedly deployed Anthropic's Claude artificial intelligence during the operation that captured former Venezuelan President Nicolas Maduro earlier this year. According to a Wall Street Journal report from February 13, the AI model was used in the mission that resulted in Maduro's apprehension and transfer to New York to face drug-trafficking charges in January.
Claude's military application came through Anthropic's existing partnership with data analytics firm Palantir Technologies. Palantir's platforms have been widely adopted by the Defense Department and federal law enforcement agencies for various operational purposes.
This deployment arrangement allowed the military to access Claude's capabilities while operating within the constraints of existing government technology infrastructure.
The reported use of Claude in a military operation directly conflicts with Anthropic's established usage policies. Company guidelines explicitly prohibit using the AI system to support violence, design weapons, or conduct surveillance activities.
These restrictions are part of Anthropic's $30 billion funding round terms, with the company currently valued at approximately $380 billion.
The Pentagon has been actively pursuing expanded AI capabilities for classified military networks. Defense officials have reportedly been pushing leading AI developers, including both Anthropic and OpenAI, to provide their tools for use on secure government systems with fewer operational restrictions than those applied to commercial users.
Anthropic currently represents the only major AI provider accessible in classified government environments, though this access comes through third-party arrangements rather than direct deployment. Even through these indirect channels, government users remain technically bound by the company's published usage policies regarding prohibited applications.
This military application follows a broader pattern of government adoption of Anthropic's technology. In August 2025, the General Services Administration established a OneGov agreement with Anthropic that provides Claude AI access to all federal government branches for a nominal $1 per agency fee.
The arrangement supports FedRAMP High workloads and includes continuous updates as new capabilities become available.
The U.S. AI Safety Institute had previously established formal collaboration agreements with Anthropic in August 2024, focusing on safety research, testing, and evaluation protocols. These agreements were designed to enable government assessment of AI capabilities and potential safety risks before widespread deployment.
Anthropic CEO Dario Amodei has publicly expressed concerns about AI's potential impacts, warning in January 2026 that artificial intelligence "will test us as a species". His 38-page essay cautioned about the technology's capacity to disrupt employment markets and create national security challenges while acknowledging its transformative potential.
The company's infrastructure strategy involves multi-cloud deployment across Google's Tensor Processing Units, Amazon's Trainium chips, and Nvidia's GPUs. This diversified approach allows workload optimization across different hardware platforms for training, inference, and research applications.
Reuters noted it could not immediately verify the Wall Street Journal's report about Claude's military deployment. The Defense Department, White House, Anthropic, and Palantir did not provide immediate responses to requests for comment regarding the reported Venezuela operation usage.
