When AI is regulated less than refrigerators, the problem is alarmingly serious | Richard Hermer

Former UK Attorney General Richard Hermer writes a warning that frontier AI is subject to less regulation than refrigerators, posing an existential risk, and calls for immediate strict regulation and stronger international coordination. He divides risks into existential risks and existential challenges: the former could allow AI to kill billions of people, for example by designing bioweapons; the latter could deprive billions of people of substantial autonomy over their lives. He says the issue only gained attention after Hugging Face was hacked and Jacob Coxon resigned from Anthropic, and argues that the future of AI cannot be left to the US, China, and big tech companies, nor can it rely on industry self-regulation.