Pentagon: Anthropic, an Unprecedented Supply Risk

The Pentagon Targets Anthropic: An Unprecedented Supply Chain Risk

The Department of Defense (DOD) has recently informed Anthropic that the company is now considered a supply chain risk. This information, reported by Bloomberg, comes from a senior official in the department, highlighting the seriousness of the situation.

This designation follows an ongoing conflict between Anthropic and the DOD. Anthropic's CEO, Dario Amodei, has firmly refused to allow the military to use its artificial intelligence systems for activities such as mass surveillance of American citizens or for the development of autonomous weapons that could make firing decisions without human intervention. The DOD, for its part, has expressed disagreement with the idea that a private contractor could limit the use of AI in military contexts.

Typically, supply chain risk designations are reserved for foreign entities considered adversaries. This new label now requires any company or agency collaborating with the Pentagon to certify that it does not use models developed by Anthropic.

Consequences for Anthropic and the Pentagon

This decision by the Pentagon could have significant repercussions for Anthropic and its operations. Until now, Anthropic was the only AI lab capable of providing systems suitable for classified missions. Currently, the U.S. military is using Claude, an Anthropic system, in its operations in Iran to quickly process operational data. Claude is integrated into Palantir's Maven Smart System, a crucial tool for military operations in the Middle East, according to Bloomberg.

The designation of Anthropic as a supply chain risk, due to this disagreement, is seen by some as an unprecedented action by the DOD. Dean Ball, former AI advisor at the White House under the Trump administration, described this decision as the "last gasp" of the American republic. He criticized an approach he describes as "tribalist," where domestic innovators are treated more harshly than foreign adversaries.

Reactions in the Tech Sector

In response to this situation, hundreds of employees from OpenAI and Google have called on the DOD to reverse its decision. They have also urged Congress to oppose what they see as an abuse of power against an American tech company. These employees have encouraged their leaders to continue refusing DOD requests regarding the use of their AI models for mass surveillance and for the development of autonomous weapons.

In this tense context, OpenAI has signed an agreement with the DOD, allowing the military to use its AI systems for "all legal uses." However, some OpenAI employees have expressed concerns about the vague wording of this agreement, fearing it could lead to the same controversial uses that Anthropic sought to avoid.

The Political Tensions in the Background

Dario Amodei described the DOD's actions as "reactionary and punitive." He also suggested that his refusal to financially support Donald Trump may have exacerbated the conflict with the Pentagon. In contrast, OpenAI's president, Greg Brockman, recently made a substantial donation of $25 million to the MAGA Inc. Super PAC, showing clear support for Trump.