Some investors are expressing growing concerns about CEO Dario Amodei as Lockheed Martin takes a firm stance against the Trump administration, even as it pledges to follow government restrictions on AI startups.
Privately, investors have voiced frustrations that Amodei has not fostered good relations and has come across as combative with Pentagon officials amid rising tensions.
Supporters of Anthropic are urging the company to navigate this situation cautiously, worried that Amodei’s approach might worsen conflicts and pose risks to the business due to its ongoing clash with the Pentagon, according to reports.
One individual familiar with the situation described it as a matter of “ego and diplomacy.”
Earlier in the week, it came to light that a controversial blog post by in-house researcher Amanda Askell resurfaced after Trump labeled the AI startup as “woke” and “radical leftist,” announcing a ban on its services for federal agencies.
Some Washington officials are uneasy about the potential impact of the company’s political leanings, especially after Askell’s posts compared meat-eating to “ritual cannibalism” and criticized the prison system.
The ongoing disagreement centers around Anthropic’s refusal to remove safeguards that prevent its Claude AI from being utilized for autonomous weapons or domestic mass surveillance.
This escalating situation is already creating ripples in the defense industry, with Lockheed Martin stating it will adhere to government directives and plan to phase out Anthropic’s technology, a move that other defense contractors are likely to follow once the startup is deemed a “supply chain risk.”
The company conveyed, “We follow the direction of the President and the Department of the Army,” indicating they believe the consequences will be minimal since they don’t depend on a single AI provider.
Lawyers close to government contractors have predicted that other defense firms will align with Lockheed’s strategy moving forward.
As Franklin Turner, a lawyer specializing in government contracts, pointed out, companies with substantial business ties to the government are already working to sanitize their supply chains due to the potential threat posed by Anthropic.
“Regardless of the legal excuses, the threat is significant… It has already caused notable harm to the company,” Turner mentioned, referring to Anthropic.
A spokesperson for L3Harris chose not to comment.
The Post has also reached out to Anthropic, General Dynamics, and RTX for their insights.
Anthropic, which has backing from tech giants like Amazon, Google, Microsoft, and Nvidia—as well as venture firms like Lightspeed, Iconiq Capital, and Coatue—faces increasing scrutiny.
Recently, Amazon CEO Andy Jassy had discussions with Amodei regarding the ongoing issues, although details of those conversations remain unclear.
It was reported that Lightspeed and Iconiq are also engaging with Anthropic executives.
The dispute heightened in late February when the Pentagon mandated that AI companies accept an “all lawful use” clause, allowing their technology to be deployed by the military without exceptions.
Despite this, Anthropic upheld its safeguards, maintaining that its Claude AI cannot be used for fully autonomous weapons or mass surveillance on U.S. soil.
Last week, the conflict escalated further when the Trump administration instructed federal agencies to cease using Anthropic’s technology, with a six-month deadline for phasing it out.
Additionally, Secretary of Defense Pete Hegseth classified the startup as a potential “supply chain risk,” a label that could prevent government contractors from utilizing its products.
This precipitated a rush among agencies and contractors to comply with the new directives.
Rival company OpenAI appears to be benefiting already from this situation, having recently announced a secret agreement with the Department of Defense. Anthropic is publicly arguing against its classification as a supply chain risk, but critics are concerned that such disputes might lead AI companies to relax their safeguards to secure lucrative defense contracts.
Following the government’s directive, the State Department is now transitioning from its internal StateChat system to an OpenAI model.
As stated by a White House spokesperson, “We will not allow woke corporations on the radical left to dictate how our military operates.” They reaffirmed that Secretary Hegseth and the Department of the Army are ensuring that U.S. troops are equipped with the necessary tools to succeed, free from interference by left-leaning corporate policies.
