Artificial intelligence presents an unprecedented existential risk to humanity, according to a former attorney general who spent two years focused on national security. This danger is of an altogether different magnitude than threats such as terrorism, armed conflicts, or the climate crisis, which traditional intelligence and defence communities are better placed to counter. The risk from AI is difficult to overestimate and has a limited time window for action.
While those involved in AI's creation have long flagged its extraordinary threats, public discourse recently gained traction following the Hugging Face hack and Jacob Coxon's resignation from Anthropic. This moment is an opportunity that urgently requires increased international coordination to regulate frontier AI. There must also be meaningful public engagement to ensure AI benefits humanity and does not centralize power among a few, a task that cannot be left only to specific nations.