AI Ethics Clash: Pentagon Demands Unrestricted Military Use of Anthropic's Technology

Pentagon Pressures Anthropic to Remove AI Restrictions for Military Use​

The US Department of Defense has reportedly given Anthropic a firm deadline to open its artificial intelligence systems for unrestricted military use or risk losing its government contract. According to a person familiar with a recent meeting, Defence Secretary Pete Hegseth told Anthropic CEO Dario Amodei that the company must comply by Friday or face potential consequences.

The development places one of the leading AI firms at the center of an escalating debate over the role of artificial intelligence in national security and warfare.

Anthropic Stands Apart on Military AI Access​

Anthropic, the developer of the chatbot Claude, remains the only major AI company that has not fully supplied its technology to a new US military internal network. While competitors have moved closer to broader integration with the Pentagon, Anthropic has maintained ethical guardrails limiting how its systems can be deployed.

Amodei has repeatedly voiced concerns about unchecked government use of AI. He has warned against fully autonomous military targeting operations and AI driven mass surveillance that could monitor dissent or assess public sentiment across vast data sets.

During Tuesday’s meeting, the tone was described as cordial. However, Amodei reportedly did not shift his stance on two key restrictions. Anthropic will not permit fully autonomous targeting systems or domestic surveillance of US citizens using its models.

Pentagon Signals Potential Contract Risks​

Defense officials indicated that failure to comply could lead to Anthropic being designated a supply chain risk. They also raised the possibility of invoking the Defence Production Act, which could grant the military broader authority to access the company’s products even without its approval of specific uses.

A senior Pentagon official stated that military operations require tools without built in limitations. The official added that the Pentagon has issued only lawful orders and that the responsibility for legal usage would rest with the military.

Each of the AI defense contracts awarded last summer to Anthropic, Google, OpenAI, and Elon Musk’s xAI is valued at up to USD 200 million. Anthropic was the first AI company approved for classified military networks and works with partners such as Palantir in those environments.

Meanwhile, xAI has stated that its Grok chatbot is ready for classified settings. OpenAI recently announced that it would join the Pentagon’s secure but unclassified AI network known as GenAI, enabling service members to use a customized version of ChatGPT for certain tasks.

AI Policy Tensions Intensify Under Current Administration​

Hegseth has publicly stated that the Pentagon will not use AI models that restrict lawful military applications. In a January speech, he said that military AI systems should operate without ideological constraints and emphasized that the Pentagon’s AI systems would not be influenced by what he termed “woke” policies.

The announcement that Grok would join GenAI came shortly after the chatbot drew global attention for generating highly sexualized deepfake images without consent.

Anthropic describes itself as safety minded and said after Tuesday’s meeting that it continues to engage in good faith discussions to ensure its models can support national security missions responsibly. The company was founded in 2021 by former OpenAI executives who sought to build AI systems with stronger safeguards.

Growing Regulatory and Political Divide​

Anthropic’s cautious approach has previously aligned with efforts to introduce oversight mechanisms for advanced AI systems. The company volunteered for third party scrutiny of its models during the prior administration to address national security risks.

However, its advocacy for stricter safeguards has at times placed it at odds with policymakers. The company has publicly criticized proposals to loosen export controls on AI chips for sales to China. It has also been involved in broader debates over state level AI regulation.

Critics argue that Anthropic’s peers, including Meta, Google, and xAI, have shown greater willingness to comply with Defense Department policies covering lawful applications. Some analysts say this dynamic limits Anthropic’s bargaining power as the Pentagon accelerates its AI adoption.

Concerns over the rapid military integration of artificial intelligence are also prompting calls for congressional oversight. Legal experts warn that existing laws may not be evolving at the same pace as AI capabilities, particularly if such systems are used in surveillance or high stakes military operations.

A Defining Moment for AI and National Security​

The standoff highlights a critical inflection point for AI companies balancing commercial opportunity, national security collaboration, and ethical commitments. As the Pentagon expands its AI footprint across classified and unclassified networks, the outcome of its negotiations with Anthropic could shape the boundaries of military AI deployment in the years ahead.