Predictably, the AI giant contends that it is clearly not a supply chain risk under Congress’s controlling statute.
Anthropic, the artificial intelligence giant, has sued the Trump administration over the Defense Department’s designation of the firm as a “supply chain risk” (SCR). The lawsuit was filed in San Francisco federal district court on Monday morning. It names a slew of federal departments, agencies, and commissions as defendants, in addition to numerous administration officials in their official capacities.
(In its lawsuit, the company named the “Department of War” as the lead defendant, adopting the Trump administration’s nomenclature for the Defense Department, including in its reference to co-defendant Pete Hegseth as the “Secretary of War.” Congress, however, created the department as “the Department of Defense,” and its head as “Secretary of Defense,” by statute (see Section 111 of Title 10, U.S. Code); a new statute or amendment would be required to alter that.)
The administration’s punitive action was in response to Anthropic’s insistence that its AI tool, Claude, not be used for lethal autonomous warfare or mass surveillance of Americans. The Defense Department and U.S. intelligence agencies have been using Claude in their classified systems for over a year. The Pentagon insisted that it be permitted “all lawful use” of the technology, and slapped the designation on Anthropic when it balked.
I wrote about the administration’s action when it occurred, analogizing it to the extortionate executive orders by which President Trump has tried to put law firms he disfavors out of business unless they agreed to various adhesive conditions (including free legal work for the president’s approved causes). As I explained, AI is evolving much faster than our law can develop to regulate it consistent with constitutional and traditional limits on government. There are serious questions about the degree to which weapons should be used without human agency (i.e., about how sure we are that AI will comply with limitations), and about whether the government should be able to apply AI analytical prowess to data about Americans that the government has begun to acquire in bulk.
In announcing the SCR designation against Anthropic, Defense Secretary Pete Hegseth ordered that “[e]ffective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.” That would cripple, if not destroy, Anthropic since it would be cut off, not just from government contracts, but from commerce with any entity that deals with the government.
Congress essentially meant the SCR designation to apply to foreign entities that could corrupt American supply chains and manufacturing. Hence, I described the application of the SCR designation to Anthropic as “ridiculous,” noting that the operative provision of federal law, Section 3252 (of Title 10, U.S. Code), defines SCR as:
[T]he risk that an adversary may sabotage, maliciously introduce unwanted function, or otherwise subvert the design, integrity, manufacturing, production, distribution, installation, operation, or maintenance of a covered system so as to surveil, deny, disrupt, or otherwise degrade the function, use, or operation of such system.
I further observed:
The Trump administration has not claimed that Anthropic poses risks that could enable hostile regimes to compromise American defense systems. To the contrary, Trump and Hegseth have been pressuring Anthropic to continue in what has been close collaboration. This is not about supply chain risk; the administration is angrily lashing out because Anthropic has declined to provide it greater latitude to incorporate its AI into American defense systems.
Furthermore, Section 3252 mandates that, before an SCR designation can be made, the government must (a) make various findings, including that the designation is necessary to protect national security (which the president’s comments show it is not) and that less intrusive measures to reduce risk are unavailable; and (b) notify Congress. Does anyone believe Hegseth and other administration officials have complied with these requirements — or that they could have done so in good faith — before announcing the SCR designation?
In its lawsuit, Anthropic contends that the statute plainly does not apply to it and that the administration failed to make the required findings. For these and other reasons (for example, the abruptness of the designation and the unhinged rhetoric of Trump and Hegseth in announcing it), the company also claims the designation is arbitrary and capricious under the Administrative Procedure Act.
We’ll follow the progress of the lawsuit.