NextFin News - U.S. President Trump signaled a dramatic reversal in his administration’s stance toward Anthropic on Tuesday, stating that a deal for the Department of Defense to utilize the startup’s artificial intelligence models is now "possible." Speaking during an interview with CNBC, U.S. President Trump noted that his administration held "very good talks" with the company’s leadership at the White House last week, marking a sharp pivot from the aggressive blacklisting that has defined the relationship over the past two months.
The shift follows a period of intense friction that began in early March when the Pentagon designated Anthropic a supply chain risk. That move effectively barred defense contractors from using the company’s Claude AI models, citing national security concerns after negotiations over military applications reportedly stalled. U.S. President Trump had previously amplified this pressure via social media, ordering federal agencies to cease all use of the technology. However, the emergence of Anthropic’s "Mythos" model—described in internal documents as the most powerful AI ever developed—appears to have altered the strategic calculus in Washington.
The Mythos model has become a focal point for both national security officials and private sector investors. According to reports from Axios, the National Security Agency has already begun testing the model’s capabilities despite the official Department of Defense blacklist. Anthropic has warned that Mythos possesses advanced cybersecurity capabilities that could both identify and exploit vulnerabilities at a scale previously unseen. This dual-use potential has created a sense of urgency within the administration to secure domestic control over the technology rather than risk its isolation or adoption by adversarial interests.
The financial stakes of this rapprochement are substantial. Anthropic’s annualized run-rate revenue reached $30 billion by early April 2026, a staggering increase from the $9 billion reported at the end of 2025. This growth has fueled private market valuations as high as $800 billion, according to secondary market data. While these figures reflect intense investor optimism, some analysts remain cautious. "The current valuation assumes a near-monopoly on high-end enterprise AI that has yet to be fully realized," noted one senior researcher at a major buy-side firm, who requested anonymity to discuss private market holdings. The researcher, known for a conservative stance on AI infrastructure spending, suggested that the $800 billion figure may be more reflective of a "scarcity premium" than long-term cash flow projections.
The potential deal also faces significant internal hurdles. Anthropic was founded on principles of "Constitutional AI," and CEO Dario Amodei has historically resisted allowing the company’s models to be used for offensive military operations or mass surveillance. While U.S. President Trump described the recent White House talks as "productive and constructive," it remains unclear if Anthropic has made concessions regarding the "lethal use" clauses that originally triggered the Pentagon’s ire. Without a formal agreement on these ethical guardrails, any partnership could face renewed legal challenges or internal employee revolts similar to those seen at other tech giants in previous years.
Broader market conditions provide a volatile backdrop for these negotiations. Spot gold (XAU/USD) is currently trading at $4,777.205 per ounce, reflecting continued safe-haven demand amid geopolitical uncertainty. Meanwhile, Brent crude oil is priced at $90.57 per barrel, as energy markets weigh the impact of shifting U.S. trade policies and defense spending priorities. For Anthropic, the path to a potential October 2026 IPO likely depends on resolving its status with the Department of Defense, as a formal government contract would provide the regulatory certainty required for a public debut of this magnitude.
Explore more exclusive insights at nextfin.ai.
