Q1. 73% of Americans support mandatory safety requirements and security standards for advanced AI models. Support is nearly identical across party lines (74% D, 74% R).
Support/OpposeAIRegulation
Q2. Only a small fraction (7%) favor an approach where regulation falls solely on users, not developers.
Support/OpposeAIRegulation
Q3. When faced with a choice between banning more powerful AI or no regulation, a plurality (48%) prefer a ban vs only 18% who choose no regulation. Bipartisan (47% D, 48% R).