Regulating AI: Where U.S. State Policy and HCI (Mis)align

State policymakers focus narrowly on technical risks while overlooking broader social impacts: The analysis of 18 state AI committee reports found that policymakers' discussions of AI risks often miss the socio-technical concerns emphasized by HCI scholars—meaning they concentrate on technical problems like bias and security while neglecting systemic issues like workforce displacement, community impacts, and power imbalances.

"Responsible AI" language masks incomplete risk assessment: Although state committees frequently invoke the concept of "responsible AI," their actual framings omit many of the risks identified in established academic taxonomies, suggesting that the term may create a false sense of comprehensive risk management without addressing the full scope of potential harms.

Federal policy vacuum is driving inconsistent state-level responses: With scant federal AI guidelines, U.S. states have independently formed AI committees to study societal trade-offs, resulting in fragmented and potentially misaligned policy approaches across different jurisdictions rather than coordinated national standards.

HCI research and policy development are disconnected: The paper identifies significant gaps between what HCI scholars understand about AI's socio-technical impacts and what state policymakers are actually considering, indicating that valuable research insights are not effectively informing policy decisions.

Participatory design and community input are largely absent from current policy processes: The authors recommend that policymakers adopt HCI approaches like participatory design to incorporate diverse stakeholder perspectives, suggesting that current state-level AI committees may not adequately represent affected communities in their deliberations.