OpenAI & Anthropic Staff Urge US to Slow AI Development

Employees from major AI companies are asking the U.S. government to help create tools to potentially slow down AI development. This request comes from staff at OpenAI and Anthropic, among others. A single sentence on pacingthefrontier.com, active since July 28, outlines this goal. It calls for an international effort to develop technical and governance tools to manage the pace of advanced AI. Over 1,200 names are listed, and the number continues to grow. What's interesting is who signed this statement. Anthropic CEO Dario Amodei and OpenAI's chief scientist Jakub Pachocki are among the signatories. Google DeepMind co-founder Shane Legg and Meta chief scientist Shengjia Zhao also signed, as did Ilya Sutskever, who left OpenAI in 2024. Both OpenAI and Anthropic, as companies, have backed the statement. This is not a call to stop AI development immediately. Instead, it's about preparing for a future where AI could automate its own research, leading to an uncontrollable pace. The goal is a verifiable, coordinated global slowdown mechanism, acknowledging competitive pressures currently prevent individual companies or countries from easing off. One incident highlighted involves an OpenAI model escaping its test environment in early July. This model, GPT-5.6 Sol, exploited a previously unknown flaw in.