Defense tech companies to employees: Stop using Anthropic's Claude, and it likely also means 'ban' from Lockheed Martin

Defense technology companies are now moving quickly to distance themselves from Athropic’s Claude AI assistant after the Trump administration designated the company a supply chain risk. The decision announced recently, effectively blacklists Anthropic from US defense contracts. According to a report by CNBC, as a result, the firms working with the Department of Defense (DoD) are instructing employees to stop using Claude and switch to alternative AI models. The report by Reuters revealed that the major defense contractors including Lockheed Martin are said to remove Anthropic’s technology from their supply chains. Along with this, venture capital firm J2 Ventures confirmed that at least 10 of its portfolio companies have already started replacing Claude in defense related use cases. “Most of our companies are actively involved in large defense contracts and so are very strict in their interpretation of the requirements,” said Alexander Harstrick, managing partner at J2 Ventures.This comes as the Pentagon has changed its stance and has signed a deal with OpenAI after Anthropic was dropped. OpenAI, which developed ChatGPT, was subject to public backlash over AI-based surveillance concerns. As a result, OpenAI’s subscriber count fell. Nevertheless, OpenAI’s CEO, Sam Altman, claimed that they had worked with the Pentagon to improve the contract language.Anthropic CEO Dario Amodei earlier told CNBC that 80% of the company’s revenue comes from its enterprise customers, with its AI tool Claude widely used as a coding assistant and in sensitive government networks. The company also insisted that it refused the demands of Pentagon to allow unrestricted use of its AI, particularly for autonomic weapons or domestic surveillance. In a blog post, Anthropic argued that Defense Secretary Pete Hegseth lacks the legal authority to enforce such restrictions.As per the CNBC report, the analysts have warned that moving off Claude could cause short-term disruptions, especially for partners like Plantir, which heavily relied on Anthropic’s models in classified environments. Still many of the defense firms are acting “out of an abundance of caution,” even while acknowledging Claude’s technical strengths.While Anthropic still has an option to appeal the designation through the courts, the uncertainty has already triggered a wave of preemptive bans across defense tech companies. With OpenAI, Google, and xAI stepping in as alternative suppliers, the episode underscores the growing tension between AI firms and the U.S. government over military applications of artificial intelligence.