The Claude neural network was used to simulate combat scenarios during the American-Israeli campaign.
Just hours after Donald Trump announced the termination of federal contracts with Anthropic and the end of use of the company's systems, the US military, according to recent reports, continued to deploy the Claude model in operations against Iran.
This refers to the joint US-Israeli strike campaign that began on Saturday. Reports of Claude's use in such an operation demonstrate how deeply generative AI tools have already become integrated into US defense processes. Quickly removing such a system from critical scenarios proves difficult if the model is already involved in planning and analysis.
According to recent reports, military structures have used Claude for intelligence analysis, target selection assistance, and combat scenario modeling related to strikes. Against this backdrop, the gap between the political decision to ban it and actual practice within the security agencies is particularly noticeable.
The latest round of conflict erupted when Trump published a post on Truth Social on Friday, hours before the expected start of the US-Israeli military campaign against Iran. In the post, Trump demanded that all federal agencies immediately cease using Anthropic's artificial intelligence products, including Claude. The president cited national security and his own authority for the decision.
Tensions between Anthropic and the US defense establishment had been growing for months. According to The Guardian , in January, the US military had already deployed Claude in an operation to capture Venezuelan President Nicolás Maduro. This scenario seemed especially controversial given Anthropic's internal regulations, which prohibit the use of Claude for violence, weapons development, or surveillance without strict restrictions.
Disagreements over individual cases gradually escalated into a broader conflict. Relations between the Pentagon, the Trump administration, and Anthropic deteriorated after the military demanded greater access to AI without ideological or corporate constraints. Defense Secretary Pete Hegseth reportedly accused the company of putting political views above national security interests and made it clear that the military would not bend to the values of tech corporations.
While defending his decision to sever ties with Anthropic, Hegseth simultaneously acknowledged that phasing out Claude isn't an immediate decision. He said the company will continue to provide services for a maximum of six months, allowing the military to transition smoothly to a new supplier. In essence, this isn't an abrupt shutdown, but a phased replacement of the existing technology.
Competitors quickly filled the vacant space. OpenAI CEO Sam Altman announced an agreement with the Pentagon to deploy the company's tools, including ChatGPT, within closed military infrastructure. A month earlier, Elon Musk's xAI also signed a contract allowing the Grok model to be used in secure US military environments. The agreement extends standard Department of Defense terms to Grok and paves the way for broader application of the model for legitimate military purposes.
The story surrounding Claude highlights a larger problem: once AI becomes part of reconnaissance, simulation, and strike preparation, the debate over acceptable boundaries quickly moves beyond corporate ethics and becomes a question of military infrastructure, government procurement, and political influence.
Just hours after Donald Trump announced the termination of federal contracts with Anthropic and the end of use of the company's systems, the US military, according to recent reports, continued to deploy the Claude model in operations against Iran.
This refers to the joint US-Israeli strike campaign that began on Saturday. Reports of Claude's use in such an operation demonstrate how deeply generative AI tools have already become integrated into US defense processes. Quickly removing such a system from critical scenarios proves difficult if the model is already involved in planning and analysis.
According to recent reports, military structures have used Claude for intelligence analysis, target selection assistance, and combat scenario modeling related to strikes. Against this backdrop, the gap between the political decision to ban it and actual practice within the security agencies is particularly noticeable.
The latest round of conflict erupted when Trump published a post on Truth Social on Friday, hours before the expected start of the US-Israeli military campaign against Iran. In the post, Trump demanded that all federal agencies immediately cease using Anthropic's artificial intelligence products, including Claude. The president cited national security and his own authority for the decision.
Tensions between Anthropic and the US defense establishment had been growing for months. According to The Guardian , in January, the US military had already deployed Claude in an operation to capture Venezuelan President Nicolás Maduro. This scenario seemed especially controversial given Anthropic's internal regulations, which prohibit the use of Claude for violence, weapons development, or surveillance without strict restrictions.
Disagreements over individual cases gradually escalated into a broader conflict. Relations between the Pentagon, the Trump administration, and Anthropic deteriorated after the military demanded greater access to AI without ideological or corporate constraints. Defense Secretary Pete Hegseth reportedly accused the company of putting political views above national security interests and made it clear that the military would not bend to the values of tech corporations.
While defending his decision to sever ties with Anthropic, Hegseth simultaneously acknowledged that phasing out Claude isn't an immediate decision. He said the company will continue to provide services for a maximum of six months, allowing the military to transition smoothly to a new supplier. In essence, this isn't an abrupt shutdown, but a phased replacement of the existing technology.
Competitors quickly filled the vacant space. OpenAI CEO Sam Altman announced an agreement with the Pentagon to deploy the company's tools, including ChatGPT, within closed military infrastructure. A month earlier, Elon Musk's xAI also signed a contract allowing the Grok model to be used in secure US military environments. The agreement extends standard Department of Defense terms to Grok and paves the way for broader application of the model for legitimate military purposes.
The story surrounding Claude highlights a larger problem: once AI becomes part of reconnaissance, simulation, and strike preparation, the debate over acceptable boundaries quickly moves beyond corporate ethics and becomes a question of military infrastructure, government procurement, and political influence.
