OpenAI, Anthropic, and Google have proposed a slowdown in frontier AI development, citing safety concerns. Anthropic CEO Dario Amodei requested antitrust exemptions to coordinate this effort, a move supported by industry figures such as Sam Altman and Elon Musk.
Aidan Gomez, CEO of Cohere, criticized the proposal as a market-locking tactic. He argues that the plan would allow dominant labs to set industry standards and create high barriers to entry, such as requirements for massive compute and permanent monitoring infrastructure.
Niels Rogge of Hugging Face dismissed the safety warnings as a strategy to restrict open-weight models and protect competitive advantages against rivals like DeepSeek. He noted that open-source models have proven effective in cybersecurity defense.
Venture capitalist David Sacks described OpenAI and Anthropic as a duopoly in frontier intelligence. He suggested that their push for a slowdown is motivated by product liability risks and the need for model reliability rather than purely altruistic safety goals.
Political reactions remain divided. President Donald Trump labeled the existential risks of AI a hoax and stated that China will not participate in slowdown agreements. Conversely, figures including Barack Obama, Bernie Sanders, and Kamala Harris support the proposed slowdown.
