“We need to take this seriously”: Hugging Face co-founder warns over the risks of AI agents

Thomas Wolf, co-founder of Hugging Face, says the argument over whether we “anthropomorphise” artificial intelligence risks obscuring a more urgent concern: as AI agents gain autonomy, cybersecurity and AI safety are rapidly becoming inseparable.

For Thomas Wolf, co-founder of Hugging Face, the debate over whether artificial-intelligence agents should be described in human terms may be missing the larger point. What matters, he argues, is whether the language helps society grasp the seriousness of the risks now emerging.

Speaking to the ETN Show, Wolf said anthropomorphising AI systems could even prove useful if it encourages people to treat their behaviour with greater caution. “I think it’s better to take this really seriously,” he said.

The rest of this piece, every live blog in full, and the documents behind the reporting. Cancel in two clicks, as always.