OpenAI and Anthropic staff circulate letter asking Washington to help pace AI development

5:45 AM · July 28, 2026

Bloomberg reported on July 28, 2026 that employees at OpenAI and Anthropic are circulating a letter addressed to the US government asking it to support a mechanism that would let the pace of frontier AI development be deliberately slowed if needed, rather than continuing to advance as fast as compute and competition allow. The document, shared among staff at both companies rather than issued by leadership, warns there is ‘a real risk’ that AI systems progress faster than people can ‘understand or control,’ pointing specifically to the growing use of AI to automate AI research itself as a mechanism that could compress development timelines further. The letter calls for an international coordination effort, rather than unilateral action by any single company or government, reflecting a recurring worry among safety minded staff that competitive pressure between labs and between the US and China makes any single company's restraint self defeating. It follows earlier employee actions at both companies, including a February 2026 letter in which OpenAI and Google staff backed Anthropic's own dispute with the US government, suggesting a pattern of rank and file researchers pushing for guardrails their employers have been reluctant to adopt unilaterally. The letter carries no legal or binding weight, but its timing, arriving the same week as the Open Secure AI Alliance launch and Anthropic's public position on open weights, underscores how unsettled the industry's internal consensus on pacing and safety has become.