Secretary of Defense Pete Hegseth is threatening unprecedented retaliation, doubtlessly labeling Anthropic as a “supply chain risk.” This designation may destroy Anthropic’s enterprise forward of their anticipated preliminary public providing (IPO). Alternatively, it’s attainable the administration may declare that Anthropic’s AI is so important to the DOD that it’ll use the Defense Production Act (DPA) to aim to compel the corporate to supply their expertise if they don’t comply with the federal government’s calls for. It must be explicitly acknowledged that these two threats are immediately in battle with one another: Either Anthropic is a danger to the DOD and must be expelled from their techniques due to that hazard or it’s so important to the DOD that our nationwide safety can be in danger with out unrestrained entry to it. It can’t be each. The Trump administration seemingly doesn’t consider both of these issues. This is a negotiating tactic to get what it needs from Anthropic.
By Friday, February 27, the DOD may basically declare struggle not on a overseas nation however on considered one of America’s most profitable frontier AI firms if it doesn’t bow to its calls for. This can be an unprecedented and pointless peacetime transfer that sends the sign to different non-public firms that they have to do the Trump administration’s bidding or face existential penalties.
Background
Anthropic has targeted on creating Claude, its proprietary AI mannequin and instruments, primarily for the enterprise software program marketplace for enterprise and government, together with offerings specific to the U.S. government’s unclassified and classified networks. Anthropic has received more than $8 billion in investment funding from Amazon and is hosted on Amazon Web Services (AWS), which incorporates a number of government-specific cloud computing services throughout classification ranges. Claude is offered on U.S. government unclassified networks and is the only frontier AI tool available to U.S. government users for use with information classified up to the secret level. While the precise authorities contracts haven’t been made public, Anthropic will need to have included some sort of phrases of service and usage policies within the DOD and the General Services Administration’s (GSA) contracts or the Pentagon wouldn’t be attempting so strongly to renegotiate that coverage.
According to Axios, “Anthropic and the Pentagon have held months of contentious negotiations over the terms under which the military can use Claude.” The situation got here to a head when it was reported that the DOD used Claude in the planning of the raid to capture Venezuelan President Nicolás Maduro and concerns that it violated the Anthropic Usage Policy (although it’s not clear what a part of the present usage policy would have been violated by the raid).
The DOD is reportedly insisting “that each one AI labs make their fashions out there for ‘all lawful uses’” while “Anthropic is willing to loosen its usage restrictions” except for “the mass surveillance of Americans” and “the development of weapons that fire without human involvement,” which—it should be noted—are only a small fraction of Anthropic’s present utilization coverage. Critically, the DOD has not stated why it objects to the restriction in opposition to utilizing Claude to develop “the mass surveillance of Americans,” which might not be a authorized motion for the Trump administration. The DOD has only reiterated their position that they want the ability to use the AI for “all lawful uses.” Anthropic could also be proper to be involved by this phrasing, as numerous federal courts have repeatedly found the Trump administration’s actions are not lawful.
There is little transparency round how detection models that identify violations of usage policies for AI chatbots are operationalized. The potential to establish and reply to violations of builders’ AI mannequin accessed via an utility programming interface (API) by a deployer is much more restricted, as CAP has written previously. The monitoring of the U.S. authorities’s use of Claude, particularly on the categorised degree, is sort of definitely extraordinarily restricted, and the sensible potential for Anthropic to find and prohibit the U.S. authorities’s use of their instruments is questionable. In reality, only when the use of Claude in the Maduro raid was leaked to the press did this situation turn out to be a part of a broader firestorm.
Terms of service and utilization insurance policies as an try to limit using superior dual-use basis fashions are unlikely to succeed, particularly with authorities use, exhibiting the true want for precise legal guidelines and regulation of AI. Still, Anthropic’s commitment to its values and its try to carry on to some restricted makes use of is admirable. Setting phrases of service to your personal merchandise is meant to be authorized for companies in America. Claude is gaining large momentum with fierce competitors within the non-public sector, and being the one frontier AI instrument out there in categorised settings has made Claude enormously invaluable to the DOD, which clearly needs to maintain utilizing the instrument.
The risk of unprecedented retaliation
At a gathering on Tuesday, February 24, 2026, Secretary of Defense Hegseth reportedly demanded Anthropic CEO Dario Amodei “give the military a signed document that would grant full access to its artificial intelligence model.” If Anthropic doesn’t comply, DOD officers are reportedly contemplating declaring Anthropic a “supply chain risk” or invoking the DPA to achieve entry to Claude with out guardrails. This threatened retaliation in opposition to an American firm is unprecedented and comes at a very pivotal second in Anthropic’s enterprise.
As former Trump administration AI advisor Dean Ball has noted, if the battle between Anthropic and the DOD had been unresolvable, the conventional place can be for DOD to cancel its contract with Anthropic. Anthropic would endure each a financial loss in enterprise and the lack of the safety fame that having the DOD as a buyer brings however would live on as an organization.
But the opposite threatened retaliations in opposition to Anthropic could possibly be critically damaging, and even deadly, to their enterprise.
According to CBS News, “because officials say they aren’t sure the government can trust Anthropic at this point, the Pentagon may decide to officially designate the company as a ‘supply chain risk’ to push them out of government.” If the DOD had been to designate Anthropic a “supply chain risk,” it could be utilizing a process previously applied only to foreign companies located in foreign adversary nations that had been thought of to be safety dangers, together with Russia’s Kaspersky Lab and China’s Huawei and ZTE telecommunications. While the implementation of this might take a number of totally different kinds—and may be sophisticated by the truth that Anthropic is an American and never a overseas firm—it could finally have a big affect on the corporate.
Being designated a “supply chain risk” would seemingly imply that DOD contractors and subcontractors wouldn’t be allowed to make use of Anthropic merchandise and would wish to certify that they didn’t use Anthropic merchandise to construct their merchandise.
This comes at a vastly important second for Anthropic as its Claude Code product is booming and changing into the coding agent of alternative for a lot of major software program firms and companies. Many of these main software program firms and companies are additionally suppliers, contractors, or subcontractors for the DOD.
While not each firm will quit Claude if Anthropic is designated a “supply chain risk,” utilizing Claude will turn out to be an affirmative option to forego any future U.S. authorities enterprise, and the best factor for an organization to do to maintain its present or future authorities enterprise is to cease utilizing Anthropic’s merchandise. On Wednesday, February 25, the DOD upped the pressure on Anthropic by asking defense contracting giants Boeing and Lockheed Martin about their use of Claude related to “a potential supply chain risk declaration.”
This may devastate Anthropic for the time being its merchandise are hitting hockey-stick development. Anthropic introduced a $14 billion income run charge in 2026 and is making ready for a attainable IPO within the subsequent yr or two. Being designated a “supply chain risk” may destroy Anthropic’s enterprise momentum and potential IPO. As Dean Ball notes, “this option could be existential for Anthropic.”
Invoking the DPA can be equally unprecedented, but according to CBS News, it would be used because, “Defense officials want full control of Anthropic’s AI technology for use in its military operations.” As Axios reports, “The idea, the senior Defense official said, would be to force Anthropic to adapt its model to the Pentagon’s needs, without any safeguards.” Dean Ball minces no phrases in describing this as “the quasi-nationalization of a frontier lab.”
While it’s unclear which a part of the DPA the administration would try to make use of to compel Anthropic to supply its AI fashions, Title I of the DPA permits for the president to:
… require that efficiency below contracts or orders … which he deems mandatory or acceptable to advertise the nationwide protection … [and can] … allocate supplies, providers, and services in such method, upon such situations, and to such extent as he shall deem mandatory or acceptable to advertise the nationwide protection.
Similarly, the definitions of “critical infrastructure” and “critical technology” together with the truth that Anthropic has an existing contract with the DOD could possibly be utilized to assist justify the DPA’s invocation.
The Biden administration’s 2023 AI Executive Order tried to make the most of the DPA’s authorities for “the national defense and the protection of critical infrastructure” to require dual-use basis mannequin builders to supply the federal government with sure data, drawing significant opposition from industry and AI supporters. The Biden AI EO was later repealed by the Trump administration in early 2025. Those who criticized the Biden administration’s use of the DPA in its AI EO ought to communicate out with equal power in opposition to the Trump administration’s way more aggressive threatened use of the identical authorities (and some are).
It is not at all clear how the DPA would or could be used by the DOD in this situation, although Alan Rozenshtein posits at Lawfare that the DOD is prone to demand both “Claude Without Contractual Restrictions” or “Forced Retraining,” which might be “the government compelling Anthropic to retrain Claude—to strip the safety guardrails baked into the model’s training, not merely modify the access terms.” Anthropic is prone to challenge any invocation of the DPA in court but “comply under protest (given the DPA provides for criminal penalties for noncompliance)” and the success of the federal government in courtroom is much from assured.
Anthropic would virtually definitely search authorized reduction from the courts if the Trump administration had been to aim to declare Anthropic a “supply chain risk” or invoke the DPA. But looking for authorized reduction can take time, and the harm to Anthropic’s enterprise within the interim could possibly be vital and irreversible. At a second of fierce competitors and nearing an IPO, even a authorities motion that will get reversed by the courts may have devastating penalties for his or her enterprise, a indisputable fact that DOD is sort of definitely conscious of and is trying to make use of to their benefit.
Additionally, we all know that when the Trump administration decides an establishment is an enemy, it may well unleash quite a few assaults on a number of fronts, as this administration’s war against Harvard University has shown. If the U.S. authorities decides Anthropic is an enemy—once more, a non-public U.S. firm that is likely one of the most profitable younger firms in American historical past—then it has quite a few levers to make its life depressing.
The Trump administration is attempting to make an instance of Anthropic
Anthropic has been labeled “woke AI” by Secretary Hegseth and the Trump administration. Anthropic is an AI company founded around concerns for AI safety, whose constitution attempts to embrace and encode certain values into AI. Anthropic has opposed among the most excessive AI deregulation—including opposing the state AI moratorium within the One Big Beautiful Bill—and is funding a pro-AI regulation super PAC. Anthropic will not be an ideal firm: For instance, it not too long ago revised its main Responsible Scaling Policy to back off its previous commitments. It can be straightforward to attempt to dismiss considerations for Anthropic’s battle with DOD as ideological and solely against the Trump administration.
However, whether or not Anthropic shares one’s values is irrelevant. This is an unprecedented try by DOD to threaten a profitable non-public American firm into bowing to its requests or face monetary destruction or nationalization. Government coercion and threats of this nature towards any U.S. AI firm must be opposed. And that is everywhere in the firm’s stance that what it believes is probably the most highly effective expertise within the historical past of the world shouldn’t be used to construct mass surveillance instruments—and the Trump administration’s clear refusal of that request. This is one other instance of the flagrant disregard for the regulation within the lengthy line of abuses from the Trump administration.
Congress ought to increase hell in regards to the administration’s try to abuse its energy and threaten to destroy considered one of America’s latest and most dear firms. The Senate and House Armed Services Committees ought to instantly convene hearings on the DOD’s threatened use of provide chain danger designations and the DPA in opposition to a home AI firm. Members of each events ought to demand that the DOD present Congress with the complete phrases of its present contracts with Anthropic and a authorized justification for any threatened retaliation. Congress ought to acknowledge that the necessity is actual for true AI safeguards and prohibited makes use of—not simply from non-public firms however codified in regulation for presidency use of AI as nicely.
There could also be some sort of de-escalation or detente between Anthropic and the Pentagon. But the harm was performed the second the federal authorities threatened to destroy or conscript an American AI firm. This administration additionally has an extended historical past of ignoring legal guidelines; pushing for the sales of companies to supporters; taking golden shares of companies they consider critical; and extra. American AI firms ought to worry what this administration threatened and the place this would possibly finish. The Trump administration has made it clear that their place is considered one of American AI dominance. However, it seems that stated dominance contains dominance over American AI firms too.
