“AI labs' consistent warnings about AI risks since ChatGPT's release initially desensitized policymakers, but the government is now reacting strongly as model capabilities have finally caught up to the labs' messaging.”