Big Tech group backs Anthropic in Pentagon AI dispute

The Information Technology Industry Council, whose members include Amazon and Nvidia, expressed concern about reports that the department might label a company a “supply-chain risk.”

The letter did not name Anthropic.

Dario Amodei, Anthropic CEO, discussed the matter with major investors and partners, including Andy Jassy, Amazon CEO.

Venture firms Lightspeed and Iconiq coordinated with others on possible solutions.

The Defense Department urged AI firms to accept an all-lawful-use clause. Anthropic said it would not allow its Claude model for autonomous weapons or mass US surveillance.

It said it would challenge any designation in court.

Defense Secretary Pete Hegseth said a designation would require all government contractors to stop using Anthropic’s technology.

Several US agencies, including the State Department, started switching to rivals like OpenAI after President Trump ordered agencies to drop Anthropic within six months.

🔗 Source: Reuters

🧠 Food for thought

Implications, context, and why it matters.

The Pentagon’s stated legal threat appears to overstate its authority

The Defense Department says contractors must stop using Anthropic’s technology based on a supply chain risk label, likely tied to 10 U.S.C. 3252 1, 2.
The statute aims at threats such as sabotage by foreign adversaries, rather than procurement disputes with a US company 1, 3.
Legal experts describe the power as narrow. It applies to purchases for specific national security systems, and it does not let the department ban a contractor’s broader commercial work 1, 2.
The Pentagon also offered a transition period after President Trump told agencies to drop Anthropic within six months. That timing weakens claims of an immediate, genuine security risk 3, 2.

The conflict is accelerating a strategic shift away from AI vendor lock-in

Several US agencies, including the State Department, are moving to rivals such as OpenAI after the Trump order. The larger effect appears to be a wider rethink of AI purchasing.
The episode raises the cost of depending on one AI provider, so enterprise customers plus the Defense Department are testing multiple vendors for redundancy 4.
Foundation models now function more like core infrastructure, so geopolitical pressure plus regulatory risk can matter as much as technical performance 4.
The public dispute increased Anthropic’s consumer visibility. Its Claude app rose to the most downloaded free app in the US, even as enterprise activity faced turbulence 4.