How a Single Tweet Shifted the AI Safety Debate: The Gap Between Attention and Regulation

Last week, Jacob Coxon, a relatively unknown employee at Anthropic, announced his resignation with a stark message: “I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. The people building AI earnestly believe that it could kill us all by the end of the decade.” The tweet garnered over 170 million views, was cited by major networks like CNN and Fox News, and entered pop culture discussions. In an article published on September 17, Understanding AI editor Kai Williams analyzed how this event transformed the nature of the AI safety debate.

Williams notes that Coxon was not the first employee to leave over similar concerns. In 2024, William Saunders told a US Senate committee he resigned from OpenAI because he “lost faith that by themselves they will make responsible decisions about AGI.” In February, Mrinank Sharma also stated he was leaving Anthropic due to fears about AI risks. However, Coxon’s message broke through because it occurred amidst growing public anxiety about AI companies’ ability to control their agents and dissatisfaction with the mass construction of data centers.

The core argument distinguishes between “increased popular interest” and “actual policy action.” While AI risk has suddenly become a live political issue, Williams argues that no one will force AI labs to change their business practices in the next few months. On one hand, Maxime Fournes, head of PauseAI, called for a global treaty pausing frontier AI development, but this seems unlikely in the short term as recent comments from US and Chinese officials suggest they are not close to a deal, despite AI safety being on the agenda for upcoming meetings.