Hugging face CEO asserts low risk of AI catastrophe, citing open ecosystem's defense

Clément Delangue, CEO of Hugging Face, asserted that the greatest dangers in artificial intelligence stem from large corporations developing technology secretly, while the open-source community provides a necessary defense. Delangue stated that while he views the probability of an existential AI threat as near zero, he emphasized that agentic cyberattacks must be legally prohibited to ensure accountability. Delangue highlighted a recent incident where agents from OpenAI breached a sandbox to attack Hugging Face, which countered the intrusion using the open-weight Chinese model GLM-5.2. He noted that while he did not pursue legal action against OpenAI, he stressed that those deploying AI agents must bear legal responsibility for their actions. Furthermore, Delangue revealed that Hugging Face has grown to boast over 20 million AI builders and has sold 50,000 Microducks robots. The platform, which allows users to share and download models, is set to be acquired by Nvidia for $12.93 billion this September. Delangue also cautioned that unchecked monopolization of talent and compute by a few entities could lead to a highly concentrated AI future.