OpenAI, Google Staff Unite Behind Anthropic's Pentagon Suit

A recent amicus brief has intensified the ongoing dialogue between Anthropic and the Department of Defense, marking a significant turn in their contentious negotiations.

The development follows Anthropic's lawsuit filed on Monday against the Department of Defense, challenging its designation as a supply chain risk. Just hours after Anthropic initiated legal action, a collective of nearly 40 employees from prominent AI companies, including OpenAI and Google — notably featuring Jeff Dean, Google’s chief scientist and lead for Gemini — submitted an amicus brief. This brief expressed their support for Anthropic’s stance, articulating their concerns regarding the Trump administration’s decision and the broader implications and risks associated with the technology.

This situation caps a tumultuous period for Anthropic, which began when the Trump administration labeled the company a supply chain risk. This classification is typically reserved for foreign entities deemed a potential national security threat. The designation came after Anthropic maintained firm "red lines" concerning the acceptable military use of its technology: explicitly prohibiting domestic mass surveillance and fully autonomous weapons (defined as AI systems capable of lethal action without human intervention). Following the breakdown of negotiations, the dispute escalated with public criticisms, and other AI companies subsequently agreed to contracts permitting "any lawful use" of their technology.

The supply chain risk designation carries severe consequences, not only barring Anthropic from securing military contracts but also blacklisting any other companies that integrate Anthropic products into their work for the Pentagon. This forces such companies to remove Anthropic's Claude AI if they wish to retain their lucrative defense contracts. Paradoxically, Anthropic’s tools, being the first AI model cleared for classified intelligence, are already deeply embedded within the Pentagon's operations. This integration was starkly highlighted when, reportedly, the U.S. military utilized Claude in the campaign that resulted in the death of Iran’s leader, Ayatollah Ali Khamenei, merely hours after Defense Secretary Pete Hegseth announced the designation.

The amicus brief asserts that Anthropic’s supply chain risk designation "is improper retaliation that harms the public interest" and underscores that the concerns underpinning Anthropic’s red lines "are real and require a response." It further advocates for a reevaluation of Anthropic’s two core prohibitions, stating that "mass domestic surveillance powered by AI poses profound risks to democratic governance — even in responsible hands" and that "fully autonomous lethal weapons systems present risks that must also be addressed."

The signatories of the amicus brief identified themselves as "engineers, researchers, scientists, and other professionals employed at U.S. frontier artificial intelligence laboratories."

They elaborated on their expertise, stating, “We build, train, and study the large-scale AI systems that serve a wide range of users and deployments, including in the consequential domains of national security, law enforcement, and military operations.” The group clarified their intent, adding, “We submit this brief not as spokespeople for any single company, but in our individual capacities as professionals with direct knowledge of what these systems can and cannot do, and what is at stake when their deployment outpaces the legal and ethical frameworks designed to govern them.”

Addressing domestic mass surveillance, the group highlighted that while data on American citizens — from surveillance cameras and geolocation to social media posts and financial transactions — is ubiquitous, "what does not yet exist is the AI layer that transforms this sprawling, fragmented data landscape into a unified, real-time surveillance apparatus." They noted that these data streams are currently siloed, but if AI were to connect them, it could combine "face recognition data with location history, transaction records, social graphs, and behavioral patterns across hundreds of millions of people simultaneously."

Regarding lethal autonomous weapons, the group expressed concerns about their unreliability in novel or ambiguous conditions that diverge from their training environments. This means they "cannot be trusted to identify targets with perfect accuracy, and they are incapable of making the subtle contextual tradeoffs between achieving an objective and accounting for collateral effects that a human can." Furthermore, the potential for AI systems to "hallucinate" necessitates human involvement in the decision-making process "before a lethal munition is launched at a human target," particularly given that the system's reasoning often remains opaque to operators and even to its developers.

In conclusion, the group behind the amicus brief affirmed, “We are diverse in our politics and philosophies, but we are united in the conviction that today’s frontier AI systems present risks when deployed to enable domestic mass surveillance or the operation of autonomous lethal weapons systems without human oversight, and that those risks require some kind of guardrails, whether via technical safeguards or usage restrictions.”