Trump orders federal agencies to stop using Anthropic products due to the US military's request for "unrestricted use" of AI leading to conflict.

President Trump has ordered federal government agencies to stop using products from the artificial intelligence company Anthropic, escalating the dispute between the AI company and the Pentagon over restrictions on technology use. On Friday, Trump stated on social media that he had instructed "all federal agencies to immediately stop using Anthropic technology," and warned that if the company does not cooperate with the transition arrangements, it may face "significant civil and criminal consequences." He also announced that related agencies, including the Department of Defense, will have a six-month "transition period" to gradually phase out the use of Anthropic products. The core of the dispute lies in the Pentagon's requirements for the use of AI models. Secretary of Defense Haggis had previously requested that the Department of Defense be allowed to use Anthropic's Claude chat Siasun Robot & Automation in any necessary way within the legal framework, without being subject to the company's terms of service restrictions. However, Anthropic insists that its models cannot be used for mass surveillance of US citizens or fully autonomous weapons systems. Anthropic has always emphasized the concept of "responsible AI" since its establishment, striving to avoid catastrophic risks brought about by technology. This stance has directly conflicted with the views of CEO Amodei and senior officials at the Department of Defense. Not long ago, the Pentagon released an artificial intelligence strategy, proposing to build an "AI-first" military and emphasizing the adoption of models that are "not subject to usage policy restrictions." Trump's move is expected to shake up the AI industry. Anthropic's contracts with the military have previously reached up to $200 million, and the company also provides services to civil agencies such as the State Department. If cooperation is completely terminated, it will weaken the company's position in the federal market. Currently valued at approximately $380 billion, there is widespread speculation that Anthropic could potentially launch its initial public offering (IPO) as early as this year. It is worth noting that Anthropic was one of the few AI systems able to operate in the Pentagon's classified cloud environment, and its "Claude Gov" tool is widely used among some defense personnel. If completely excluded from the government system, it may pose certain national security and technological substitution challenges in the short term. Meanwhile, competitors are speeding up their efforts to win government business. Musk's xAI has been approved to participate in classified projects, and OpenAI and Alphabet Inc. Class C's Gemini model are actively expanding in the federal market. OpenAI CEO Altman stated in an internal memo that the company is in communication with the Department of Defense to provide model services with certain restrictions in place. Trump's decision may also spark backlash in Silicon Valley. Previously, many tech companies, including Amazon.com, Inc. and Microsoft Corporation, had employees urging management to reject the Pentagon's proposal for "unrestricted use." Anthropic reiterated its stance on Thursday, stating, "These threats will not change our decision, as we cannot compromise our conscience to accept such demands."