Hugging Face CEO responds to former Anthropic researcher's AI extinction risk claims

On the evening of September 11, Hugging Face CEO Clément Delangue took to X to respond to comments made by former Anthropic researcher Jacob Cawson about AI extinction risk. Delangue compared Cawson's discussion to "an air-conditioning repairman talking about climate change," arguing that such views need to be viewed in a fuller context and that voices from across the AI ecosystem should be heard.

Cawson announced his resignation from Anthropic on Tuesday, September 8. He previously worked at OpenAI and expressed concerns that both companies are "racing full speed toward self-improving superintelligence" and betting with human lives. "People building AI genuinely believe that AI could kill all of us before the end of the 2020s. This is not a marketing gimmick," Cawson said. His resignation has sparked fresh debate about AI risk within the industry.

Delangue later clarified that his "air-conditioning repairman" analogy was not meant to discredit Cawson. "It just means we should look at any single point of view in a bigger context and listen to the voices of the wider AI ecosystem," he wrote. He added that, to his knowledge, Cawson's work was mainly focused on pretraining — using large datasets to train AI models — rather than on risk-focused research.