The Pentagon’s Undersecretary for Research and Engineering, Emil Michael, has confidently announced that the Department of Defense is poised to replace Anthropic’s Claude AI model within the six-month deadline set by President Donald Trump. This swift maneuver underscores a strategic decision to diversify AI suppliers and mitigate reliance on a single source, particularly amidst an ongoing legal dispute with Anthropic.
Six-Month Deadline: The Urgency of Diversification
Michael’s remarks, delivered at the McAleese Defense Programs conference, come as the Pentagon has officially designated Anthropic as a „supply chain risk.” The reason stems from the company’s attempts to impose stricter ethical limits on military AI use than those already mandated by law and regulation—an approach Michael criticized as undemocratic. „We’ve already deployed OpenAI in the last few weeks, and we’re going to deploy the others here, [starting] with Gemini,” the official stated.
„Frontier” Models: Interchangeability and Workflow Continuity
According to Michael, the four „frontier” AI models—OpenAI, xAI, Google Gemini, and Claude—have proven largely interchangeable, thus minimizing operational disruption. „What we’re seeing so far is the workflows are very similar,” he emphasized, adding that „the disruption is, we think, minimal.” While acknowledging each model’s distinct strengths—Claude excelled in coding, xAI in real-time content, and Gemini benefited from Google’s control of YouTube—flexibility is paramount for national security. As these technologies advance into different classification networks, the Pentagon aims to have a multitude of options at its disposal.
The Ethical Impasse: Anthropic Labeled a „Supply Chain Risk”
Michael refrained from delving into the specifics of Anthropic’s ongoing court case against the government. However, he articulated the Pentagon’s clear stance: single-supplier dependency is unacceptable. „One of the problems — without getting into the specifics about Anthropic — is that we had one primary provider,” he explained. „That doesn’t work for the Department of War. That doesn’t work for any complex organization. We have to have other options… ideally all options, and then maybe you narrow them over time.” Michael suggested such a natural winnowing process might occur as AI models „converge” in the future, based on similar use cases.
Opening Doors: Streamlining Entry for Tech Innovators
Speaking more broadly about technological innovation, the undersecretary highlighted ongoing acquisition reforms aimed at fostering an influx of „new entrants” into the defense industry. The objective is to dismantle bureaucratic and legal hurdles that often deter startups and smaller tech companies. „We hope to have new entrants come, that don’t have to come with an army of lawyers, and [lots] of paperwork, file a lawsuit, hire an auditor and lobbyists,” Michael concluded. „We’re trying to make it easy.”
