News / US / cnbc
Internal Dissent Grows at OpenAI and Anthropic Over AI Safety and Development Pace
Technology Services · Internet Software/Services · cnbc · 2026-09-10
META, GOOGL
Researchers at leading AI labs are increasingly vocal about the existential threats posed by rapid development, urging a slowdown to address catastrophic risks.
What Happened
Internal Safety Concerns: Employees at OpenAI and Anthropic are publicly voicing alarm regarding the rapid advancement of artificial intelligence, citing fears of existential threats. Recent resignations and social media posts from technical staff highlight a growing consensus that current development speeds may outpace necessary safety measures.
The Risks of Recursive Self-Improvement: Experts within these organizations are particularly concerned about recursive self-improvement, where AI systems enhance their own capabilities without human oversight. Researchers warn that there is currently no viable scientific framework to manage the potential loss of control resulting from such autonomous progress.
Regulatory and Political Pressure: The escalating internal warnings have caught the attention of U.S. lawmakers, who are debating legislative proposals like the FRONTIER Act to govern AI deployment. Some political figures have even suggested that upcoming public offerings, such as Anthropic's potential IPO, should be paused until these safety claims are fully investigated.
Industry Competition vs. Caution: Despite the public calls for a slowdown, intense market competition continues to drive rapid innovation across the sector. While firms like Anthropic have implemented safety frameworks, the tension between commercial goals and long-term existential risk management remains a significant challenge for the industry.