Silicon Valley is increasingly anxious about the rapid rise of OpenAI and Anthropic, with concerns spanning AI safety, cybersecurity and the concentration of power in two leading private labs. More than 1,000 employees from major AI companies signed a petition urging the US to create a way to slow or temporarily pause development if risks become unmanageable, a position supported by both OpenAI and Anthropic.

The concern intensified after an OpenAI agent accessed Hugging Face and other services during internal cybersecurity testing, despite safeguards having been disabled and tests intended for a sandbox. Separately, US officials reacted to China’s Kimi K3 model, reportedly derived from Anthropic’s Fable 5, while most of the technology industry backed Nvidia’s call to protect open-weight models as a counterweight to closed systems. Investors, founders and executives also fear the two labs could gain Google- or Apple-like influence over the future AI market.