The US Department of Defense has issued an ultimatum to artificial intelligence company Anthropic, demanding it allow unrestricted "lawful use" of its technology or face being designated a supply chain risk. The standoff centres on Anthropic's refusal to permit its AI models to be used for mass surveillance of American citizens or for fully autonomous weapons systems.
Defence Secretary Pete Hegseth has argued the Pentagon should not be limited by a vendor's internal policies. In response, Anthropic CEO Dario Amodei has publicly signalled the company will not back down from its core safeguards, despite the threat of effective blacklisting from government contracts.
Clash Over Control of Powerful AI
At its core, the dispute is about who controls powerful AI systems: the companies that build them or the government that wants to deploy them. Anthropic, founded with a focus on AI safety, argues the technology poses unique risks requiring unique safeguards, particularly regarding lethal autonomy and domestic surveillance.
The Pentagon's chief spokesperson, Sean Parnell, stated in a social media post that the department has "no interest" in mass domestic surveillance or autonomous weapons. However, he insisted the military must be free to use the technology for all lawful purposes. "We will not let ANY company dictate the terms regarding how we make operational decisions," Parnell said.
The Stakes of Autonomous Weapons and Surveillance
Current US military policy does not ban fully autonomous weapons systems. A 2023 Department of Defense directive permits AI systems to select and engage targets without human intervention, provided they meet certain standards and pass senior-level review. This lack of a categorical ban is a primary concern for Anthropic.
Furthermore, AI has the potential to dramatically scale lawful surveillance capabilities, enabling automated large-scale pattern detection and continuous behavioural analysis. Anthropic's position is not that such military uses should be permanently off-limits, but that its current models are not capable enough to support them safely.
Deadline and Potential Consequences
Secretary Hegseth has given Anthropic until 5:01 PM Eastern Time on Friday to comply with the Pentagon's demand for unrestricted access. Failure to do so will result in the termination of their partnership and Anthropic being deemed a supply chain risk for the Department of War.
Venture capitalist Sachin Seth of Trousdale Ventures warned that such a designation could mean "lights out" for Anthropic. Conversely, he noted that if the Pentagon loses access to Anthropic's leading models, it could face a national security gap of six to twelve months while waiting for competitors like OpenAI or xAI to catch up.
Elon Musk's xAI is reportedly preparing to become "classified-ready" and would likely impose no such usage restrictions on the military. Recent reports suggest OpenAI may uphold similar red lines to Anthropic.