Politics

Pentagon's push for unrestricted AI tools sparks clash with Anthropic over defence use restrictions

Tuesday, 17 February 2026 7:16AM UTC

A heated dispute between the Pentagon and AI firm Anthropic reveals a widening divide over the use of generative AI in defence, risking a partnership rupture amid national security debates.

A dispute between the Pentagon and Anthropic has escalated into a possible rupture of their partnership as Washington presses leading AI developers to broaden how their systems may be used in defence operations. According to reporting, the Department of Defense has sought commitments from Anthropic, OpenAI, Google and xAI to permit model use for “all lawful purposes”, a demand that Anthropic has resisted. ^[2],^[3]

The standoff carries financial weight: the Pentagon has signalled it could cancel roughly $200 million in business with Anthropic if the company will not relax its restrictions. Negotiations have reportedly stretched for months, with unnamed administration officials saying one firm agreed to the Pentagon’s terms and two others showed flexibility, while Anthropic remains the least amenable. ^[3],^[4]

Tension intensified after reports that Anthropic’s Claude model was employed in the operation to detain Venezuelan former president Nicolás Maduro, a claim that has become a flashpoint in talks. Anthropic has pushed back against assertions about specific missions, saying "we have not discussed the use of Claude for specific operations with the Department of War". The company says current discussions centre on “a specific set of Usage Policy questions , namely, our hard limits around fully autonomous weapons and mass domestic surveillance.” ^[4],^[5]

Anthropic’s stance reflects deeper unease inside the company about potential misuse of its most powerful systems. Company executives have warned publicly that advanced Claude variants could be abused to enable "heinous crimes," including aiding the development of chemical weapons, and have argued for stronger safeguards and transparency as models grow more capable. The Future of Life Institute and other groups have amplified those calls, launching campaigns urging tougher regulation. ^[2]

The Pentagon has defended its position as necessary to ensure military effectiveness. "Our nation requires that our partners be willing to help our warfighters win in any fight," Chief Pentagon spokesman Sean Parnell said, underscoring officials’ impatience with what some describe as operationally problematic limits. Some defence figures have begun to characterise Anthropic as a potential supply-chain risk, and officials are reportedly considering measures that could reduce reliance on the company. ^[7],^[6]

The dispute highlights a widening fault line between firms that prioritise built‑in usage constraints and a defence establishment seeking unfettered access for intelligence, battlefield support and weapons development so long as activities are legal. Industry and policy observers warn the contest could shape not only contracts but also future norms around the military applications of generative AI, as regulators, advocacy groups and companies wrestle with how to prevent catastrophic misuse while preserving operational capabilities. ^[2],^[6]

Source Reference Map

Inspired by headline at: ^[1]

Sources by paragraph:

Paragraph 1: ^[3],^[4]
Paragraph 2: ^[3],^[4]
Paragraph 3: ^[4],^[5]
Paragraph 4: ^[2],^[5]
Paragraph 5: ^[7],^[6]
Paragraph 6: ^[2],^[6]

Source: Noah Wire Services

More on this

https://www.techradar.com/pro/security/pentagon-may-sever-anthropic-relationship-over-ai-safeguards-claude-maker-expresses-concerns-over-hard-limits-around-fully-autonomous-weapons-and-mass-domestic-surveillance - Please view link - unable to able to access data
https://www.axios.com/2026/02/11/anthropic-claude-safety-chemical-weapons-values - Anthropic has issued a warning that its advanced AI models, Claude Opus 4.5 and 4.6, are vulnerable to misuse in the execution of 'heinous crimes,' including the development of chemical weapons. This revelation comes from a newly released sabotage report highlighting the increasing risks posed by powerful AI systems, especially when harmful actions are executed autonomously by the models themselves. Anthropic CEO Dario Amodei underscored these concerns in an early 2026 essay, emphasizing the real danger of a large-scale AI-enabled attack that could result in mass casualties. Despite confidence in the continuity and reliability of their newer models based on prior deployments, Anthropic acknowledges that future AI advancements or broader deployments could undermine current safety assumptions. Meanwhile, the Future of Life Institute has launched an $8 million ad campaign advocating for stronger AI regulations to prevent misuse. Amodei also noted the lack of incentive for tech companies to be transparent about the true risks AI poses due to financial and political interests.
https://dataconomy.com/2026/02/16/anthropic-resists-pentagon-push-for-autonomous-weapon-ai-use - The Pentagon is pressuring major artificial intelligence developers to permit U.S. military applications of their technology for 'all lawful purposes.' This demand is directed at Anthropic, OpenAI, Google, and xAI. An anonymous Trump administration official reported that one of these firms has agreed to the terms, while two others have demonstrated flexibility. Anthropic remains the most resistant entity in these negotiations, prompting the Department of Defense to threaten the termination of a $200 million contract with the company. Friction between the AI firm and government officials is not a recent development. In January, the Wall Street Journal documented significant disagreement regarding the permissible scope of Anthropic’s Claude models within defense contexts. The report further noted that the technology was utilized during the U.S. military operation to apprehend former Venezuelan President Nicolás Maduro. When contacted by TechCrunch, Anthropic did not provide an immediate response to inquiries regarding the contract dispute. Addressing the nature of the dispute, an Anthropic spokesperson told Axios that the company has 'not discussed the use of Claude for specific operations with the Department of War.' The spokesperson clarified that current discussions are 'focused on a specific set of Usage Policy questions — namely, our hard limits around fully autonomous weapons and mass domestic surveillance.' This statement distinguishes the company’s policy concerns from the Pentagon’s request for broader access to its models.
https://www.yahoo.com/news/articles/pentagon-threatens-cut-off-anthropic-022638347.html - The Pentagon is considering ending its relationship with artificial intelligence company Anthropic over its insistence on keeping some restrictions on how the U.S. military uses its models, Axios reported on Saturday, citing an administration official. The Pentagon is pushing four AI companies to let the military use their tools for 'all lawful purposes,' including in areas of weapons development, intelligence collection, and battlefield operations, but Anthropic has not agreed to those terms and the Pentagon is getting fed up after months of negotiations, according to the Axios report. The other companies included OpenAI, Google, and xAI. An Anthropic spokesperson said the company had not discussed the use of its AI model Claude for specific operations with the Pentagon. The spokesperson said conversations with the U.S. government so far had focused on a specific set of usage policy questions, including hard limits around fully autonomous weapons and mass domestic surveillance, none of which related to current operations. The Pentagon did not immediately respond to Reuters' request for comment. Anthropic's AI model Claude was used in the U.S. military's operation to capture former Venezuelan President Nicolas Maduro, with Claude deployed via Anthropic's partnership with data firm Palantir, the Wall Street Journal reported on Friday. Reuters reported on Wednesday that the Pentagon was pushing top AI companies including OpenAI and Anthropic to make their artificial intelligence tools available on classified networks without many of the standard restrictions that the companies apply to users.
https://www.axios.com/2026/02/13/anthropic-claude-maduro-raid-pentagon - Friction point: The Pentagon wants the AI giants to allow them to use their models in any scenario so long as they comply with the law. Anthropic, which has positioned itself as the safety-first AI leader, is currently negotiating with the Pentagon around its terms of use. The company wants to ensure in particular that its technology is not used for the mass surveillance of Americans or to operate fully autonomous weapons. The company is confident the military has complied in all cases with its existing usage policy, which has additional restrictions, a source familiar with those discussions told Axios. What they're saying: 'We cannot comment on whether Claude, or any other AI model, was used for any specific operation, classified or otherwise,' the Anthropic spokesperson said. 'Any use of Claude — whether in the private sector or across government — is required to comply with our Usage Policies, which govern how Claude can be deployed. We work closely with our partners to ensure compliance.'
https://www.ynetnews.com/tech-and-digital/article/bjoye2juze - AI systems are capable of processing large volumes of data and generating real-time insights, a capability viewed as critical in military operations. Anthropic, the maker of the chatbot Claude, has declined to accept the broad 'lawful purpose' definition. The company maintains restrictions in two areas: mass surveillance of U.S. citizens and the development or deployment of fully autonomous weapons. After months of inconclusive negotiations, Pentagon officials are increasingly inclined to sever ties with the company rather than negotiate case by case or accept preemptive limits on certain uses, Axios reported. The senior official cited by the publication said the administration views the two restricted areas as containing significant 'gray zones,' where it may be difficult to determine what qualifies under the definitions. 'Everything is on the table, including scaling back the partnership with Anthropic or cutting it off entirely,' the official said. An Anthropic spokesperson said the company remains 'committed to the responsible use of AI in support of U.S. national security.'
https://www.foxnews.com/politics/maduro-raid-questions-trigger-pentagon-review-top-ai-firm-potential-supply-chain-risk/ - Its AI model, Claude, was the first model brought into classified networks. Now, 'The Department of War’s relationship with Anthropic is being reviewed,' chief Pentagon spokesman Sean Parnell told Fox News Digital. 'Our nation requires that our partners be willing to help our warfighters in any fight.' According to a senior administration official, tensions escalated when Anthropic asked whether Claude was used for the raid to capture Maduro, 'which caused real concerns across the Department of War indicating that they might not approve if it was.' 'Given Anthropic’s behavior, many senior officials in the DoW are starting to view them as a supply chain risk,' said a senior War Department official. 'We may require that all our vendors and contractors certify that they don’t use any Anthropic models.' The officials did not elaborate on when Anthropic made the inquiry or to whom. Axios reported, citing a senior administration official, that Anthropic raised the question with an executive at Palantir, its partner in Pentagon contracting. Palantir could not immediately be reached for comment. Anthropic disputed that characterization. A spokesperson said the company 'has not discussed the use of Claude for specific operations with the Department of War' and has not discussed such matters with industry partners 'outside of routine discussions on strictly technical matters.' The spokesperson added that Anthropic’s conversations with the Pentagon 'have focused on a specific set of Usage Policy questions — namely, our hard limits around fully autonomous weapons and mass domestic surveillance — none of which relate to current operations.'

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 8

Notes: The article presents recent developments regarding the Pentagon's potential severance of ties with Anthropic over AI safeguards. The earliest known publication date of similar content is February 13, 2026, as reported by Axios. ([axios.com](https://www.axios.com/2026/02/13/anthropic-claude-maduro-raid-pentagon?utm_source=openai)) The article includes updated data but recycles older material, which raises concerns about freshness. Additionally, the article includes updated data but recycles older material, which raises concerns about freshness. The narrative has appeared across multiple reputable sources, including Axios and TechRadar, indicating a high level of coverage. However, the presence of similar content across multiple reputable sources suggests that the information is not entirely original. Given these factors, the freshness score is moderate.

Quotes check

Score: 7

Notes: The article includes direct quotes attributed to Anthropic spokespersons and Pentagon officials. However, these quotes cannot be independently verified through the provided sources. The absence of direct links to the original statements raises concerns about the authenticity and accuracy of the quotes. Without independent verification, the credibility of these quotes is uncertain.

Source reliability

Score: 8

Notes: The article cites reputable sources such as Axios and TechRadar, which are known for their journalistic standards. However, the reliance on a single source for key information, particularly regarding the Pentagon's internal deliberations, raises concerns about the independence and completeness of the reporting. The lack of corroboration from multiple independent sources diminishes the overall reliability of the information presented.

Plausibility check

Score: 7

Notes: The claims made in the article align with known industry trends and previous reports about the Pentagon's interest in AI technologies. However, the absence of specific details, such as direct quotes from Pentagon officials or Anthropic representatives, makes it difficult to fully assess the plausibility of the claims. The reliance on unnamed sources and the lack of direct evidence weaken the overall credibility of the narrative.

Overall assessment

Verdict (FAIL, OPEN, PASS): FAIL

Confidence (LOW, MEDIUM, HIGH): MEDIUM

Summary: The article presents information about the Pentagon's potential severance of ties with Anthropic over AI safeguards, citing reputable sources. However, the reliance on a single source for key information, the absence of independently verifiable quotes, and the lack of corroboration from multiple independent sources raise significant concerns about the accuracy and reliability of the content. Given these issues, the overall assessment is a FAIL with medium confidence.

AI ethics
Military technology
Artificial intelligence