The US legal system is struggling to keep up with AI, grappling with cases where chatbots provided counsel, generated evidence, or helped plan a mass shooting (Evan Ratliff/Bloomberg)

The U.S. legal system is confronting cases where AI chatbots have offered legal counsel, generated evidence, and assisted in planning violent attacks. Anthropic CEO Dario Amodei urges an immediate slowdown of AI development, proposing permanent third-party evaluators with employee-level access, coordinated democratic oversight, and global collaboration with authoritarian regimes; OpenAI has pledged to adopt the evaluator model. He warns that an OpenAI- or Hugging Face-style agent swarm could dominate the internet within 6–12 months.