America is in the wrong AI race with China

U.S. policymakers have framed the AI rivalry with China as a contest of technical superiority, shaping a light-touch regulatory stance based on the fear that rules will cede leadership. That approach overlooks growing public distrust and political backlash: a majority of Americans are now more worried than excited about AI, trust in government regulation is weak, and local opposition to data-center demands for water, electricity, and land is emerging. High-profile missteps and market reactions underline this gap - OpenAI’s Sora deepfake issue cost a $1 billion Disney licensing deal, Airbnb publicly uses Alibaba’s Qwen for cost and speed advantages, and Anthropic’s Fable has seen slow adoption as cheaper open-source models gain traction. Effective consumer protections are specific and practical: mandatory disclosure when users interact with AI, labeling of synthetic content, rights to challenge consequential automated decisions, and avenues for redress.

China is pursuing a different race by governing AI deployment rather than only model capabilities, instituting rules on transparency, protections for children and the elderly, and limits on algorithmic price discrimination. That approach coexists with strong domestic AI development and far higher public trust (about 87% in China versus 23% in the U.S.). Global winners will be those who secure public trust through deployment-focused rules and protections, not those who only demonstrate raw model performance or compute.