Anthropic, one of the most prominent artificial intelligence companies in the United States, is facing a deteriorating relationship with the Trump administration — and the fallout could limit its ability to shape the policies that govern its own industry.
According to Axios, sources say Anthropic has repeatedly struggled to communicate effectively with administration officials. The tensions came to a head over a cybersecurity executive order: a White House official told Axios that Anthropic "screwed us" after the company failed to "honor" the order. The report describes this as the latest in a pattern of friction, not an isolated incident.
The Economist separately characterized the situation as a White House "attack" on Anthropic, underscoring that the dispute has escalated beyond a private disagreement into something more public and pointed.
The stakes are high. AI companies are scrambling to influence how Washington regulates the technology — from safety standards to export controls to liability rules. A company that has burned bridges with the administration loses a seat at the table when those decisions get made, and Anthropic has positioned itself as a safety-focused alternative to rivals like OpenAI and Google DeepMind. Being sidelined in policy conversations would undercut that identity.
Why it matters: if one of America's leading AI safety advocates can't maintain a working relationship with the White House, the administration may craft AI policy with less input from the voices most focused on keeping the technology from going wrong.