Tesla (TSLA) CEO Elon Musk has fired back at Anthropic, accusing the AI company of stealing training data at a massive scale. He claimed the firm had paid multi-billion-dollar settlements linked to the alleged misconduct. The remarks surfaced as Anthropic faces growing pressure from U.S. defense officials over the use of its Claude AI model.
Claim 30% Off TipRanks
Trade AMZN with leverageMusk Criticizes Anthropic’s Practices
On Feb. 23, Musk tweeted on X that “Anthropic is guilty of stealing training data at massive scale,” a claim he described as “just a fact.” The post included images highlighting alleged industrial-scale distillation attacks and Anthropic’s own training practices. When an X user pushed back at Musk, noting that he didn’t personally create his training data either, the SpaceX CEO replied that at least they were “not super smug, sanctimonious and hypocritical about it like Anthropic.”
Notably, Musk’s criticism of Anthropic comes as the AI company has claimed that recent distillation attacks have targeted its models. According to Anthropic, three Chinese artificial intelligence firms, DeepSeek AI, Moonshot AI, and MiniMax (HK:0100), used roughly 24,000 fake accounts to pull over 16 million outputs from their Claude chatbot. The company described these actions as “industrial-scale campaigns.” It warned that illicit distillation could bypass safeguards built into its AI models and feed capabilities into foreign military, intelligence, and surveillance systems.
Pentagon Pressures Anthropic Over Claude AI
U.S. Defense Secretary Pete Hegseth reportedly called Anthropic CEO Dario Amodei to the Pentagon to discuss the military’s use of the company’s Claude AI model. The meeting was described as highly sensitive, as Anthropic had not agreed to lift certain limitations on its system despite requests from defense officials.
Claude is currently the only AI deployed in classified U.S. defense networks through a $200 million pilot program signed last year. In a Jan. 9 memo, Hegseth asked AI firms involved to revisit their agreements to ease restrictions. Anthropic, however, kept in place rules that limit activities such as mass surveillance and the development of fully autonomous weapons.
Officials cautioned that Anthropic’s refusal could flag the company as a potential supply chain risk, potentially invalidating contracts and restricting other defense partners from using Claude. Meanwhile, the Pentagon has agreements with Musk’s xAI and is close to a deal with Alphabet Class A (GOOGL) for its Gemini AI model. In addition, Nvidia (NVDA) remains a key supplier of advanced AI chips that enable the deployment of these systems in defense networks.
Anthropic: The Private AI Firm Behind Claude
Founded in 2021, Anthropic is a privately held AI company. It has revealed plans to go public in 2026, potentially intensifying competition with firms such as Google and Nvidia. It’s an AI model. Claude is a next-generation assistant designed for secure, reliable enterprise use. Claude is designed to handle tasks such as document review, data visualization, and SQL reporting, underscoring Anthropic’s focus on scalable, safety-conscious AI solutions.
What are the Best AI Stocks to Buy Now?
According to TipRanks stocks center, some of the top AI stocks investors are watching today include GOOGL, NVDA, TSLA, Amazon (AMZN), Apple (APPL), and others.
