Rep. Khanna pushes for US-China AI safety treaty

Rep. Ro Khanna sent letters to ODNI and Chinese AI firms DeepSeek, Alibaba and Moonshot AI asking about safeguards against catastrophic AI risks. He seeks a US-China treaty banning recursive self-improvement and monitoring frontier labs, though he admits it is unlikely soon.

• Khanna asks ODNI to assess US readiness if an AI lab loses control of its bots
• Letters to DeepSeek, Alibaba and Moonshot AI seek safeguards and kill-switch details
• He proposes a treaty banning recursive self-improvement and requiring lab inspections
• Khanna called Trump's meeting with tech CEOs a 'branding exercise'