The U.S. Army employed Claude from Anthropic in an operation to capture Venezuelan President Nicolás Maduro, according to sources cited by the WSJ.

The mission involved bombing several locations in Caracas.

Using the model for such purposes contradicts Anthropic's public policy, which explicitly prohibits the application of AI for violence, weapon development, or surveillance organization.

“We cannot comment on whether Claude or any other model was used in a specific operation—secret or otherwise. Any application of LLM—both in the private sector and government entities—must comply with our policy governing the deployment of neural networks. We work closely with partners to ensure compliance,” said an Anthropic representative.

The integration of Claude into the Department of Defense was made possible through Anthropic's partnership with Palantir Technologies, whose software is widely used by military and federal law enforcement agencies.

After the raid, an Anthropic employee inquired with a colleague at Palantir about the specific role the neural network played in the operation to capture Maduro, according to the WSJ. A representative from the startup stated that the company had not discussed the use of its models in specific missions “with any partners, including Palantir,” limiting discussions to technical matters.

“Anthropic is committed to using advanced AI to support U.S. national security,” the company representative added.

Anthropic vs. Pentagon?

Pentagon spokesperson Sean Parnell announced a review of relations with the AI lab.

“Our country needs partners willing to help our fighters win any war,” he stated.

In July 2025, the U.S. Department of Defense awarded contracts worth up to $200 million to Anthropic, Google, OpenAI, and xAI for the development of AI solutions in security. The department's Digital and AI Technologies Directorate planned to use their developments to create security agent systems.

However, by January 2026, the WSJ reported on the risk of terminating the agreement with Anthropic due to disagreements over the startup's strict ethical policy. The rules prohibit the use of the Claude model for mass surveillance and autonomous lethal operations, limiting its application by agencies like ICE and the FBI.

Officials' dissatisfaction grew amid the integration of the Grok chatbot into the Pentagon's network. Defense Secretary Pete Hegseth emphasized that the department “will not use models that do not allow for warfare” when commenting on the partnership with xAI.

Pressure on Developers

Axios reported, citing sources, that the Pentagon is pressuring four major AI companies to allow the U.S. military to use their technologies for “all lawful purposes,” including weapon development, intelligence gathering, and combat operations.

Anthropic refuses to lift restrictions on surveillance of U.S. citizens and the creation of fully autonomous weaponry. Negotiations have stalled, but quickly replacing Claude is challenging due to the model's technological superiority in specific government tasks.

In addition to Anthropic's chatbot, the Pentagon uses OpenAI's ChatGPT, Google's Gemini, and xAI's Grok for non-classified tasks. All three have agreed to relax restrictions that apply to regular users.

Currently, discussions are underway to transfer LLMs to a classified environment and their use “for all lawful purposes.” One of the three companies has already agreed to this, while the other two are “showing greater flexibility” compared to Anthropic.

Militarization of AI

The U.S. is not the only country actively integrating artificial intelligence into its defense sector.

China

In June 2024, China developed an AI commander for large-scale military simulations involving all branches of the PLA. The virtual strategist is endowed with extensive powers, learns quickly, and improves tactics during digital exercises.

In November, media reported that Chinese researchers adapted Meta's Llama 13B model to create the ChatBIT tool. The neural network was optimized for gathering and analyzing intelligence data, as well as supporting operational decision-making.

India

New Delhi has also bet on AI as a driver of national security. The government has developed national-level strategies and programs, established special institutes and bodies for AI implementation, and launched projects to apply the technology across various sectors.

United Kingdom

London has designated artificial intelligence as a priority area. In the “AI Strategy for Defense” (2022), the department views AI as a key component of future armed forces. In the “Strategic Defense Review” (2025), the technology is identified as a fundamental element of modern warfare.

Whereas AI was previously seen as an auxiliary tool in military contexts, the British Armed Forces now plan to transform into “technologically integrated forces,” where AI systems will be applied at all levels—from command analytics to the battlefield.

Recall that in March 2025, the Pentagon announced the use of AI agents for modeling confrontations with foreign adversaries.