Kevin Frazier (The University of Texas School of Law) has posted Will AI Pass the Kayla Test? on SSRN. Here is the abstract:
Kayla drives for Lyft in Morgantown, West Virginia. Asked about AI, she said nothing about chatbots. She spoke about data centers—about her family’s land, her children’s job prospects, and a state asked once again to bear the costs of a national economic project whose benefits accrue somewhere else. Her story rarely enters AI policy debates.
This essay argues it belongs at their center. The argument proceeds from a simple observation. West Virginia has among the lowest rates of AI adoption in the nation and among the fastest-growing concentrations of AI infrastructure. That mismatch is not an accident. It is the predictable result of a technology developed on the coasts and powered in between. If the land, water, and political tolerance of working-class communities are going to sustain a national project, then those communities must be made whole—and given a genuine voice in the project’s direction.
The essay pursues that goal in three parts. Part I proposes the terms of a new social bargain: a national commission that converts data center siting from a local imposition into a compensated national project; an AI Literacy Corps that pairs underemployed young Americans with the communities least served by AI; and a democratically ratified process for selecting the values that steer frontier models. Part II supplies the informational precondition for that bargain—a standardized AI Values Framework disclosure regime, administered by the states and grounded in two forgotten chapters of American regulatory history: corporate purpose clauses and food nutrition labels. Part III explains why none of this can wait. Public frustration with AI is mounting faster than existing institutions can process it, and history teaches that such pressure demands release. The essay accordingly proposes a Commission on AI, modeled on the United States Industrial Commission of 1898, through which Congress can channel that frustration into considered legislation rather than blunt reaction.
The labs are not being asked to sacrifice. They are being offered a trade: concrete obligations in exchange for legitimacy, regulatory predictability, and a public with genuine reasons to want AI to succeed. The question is whether the industry accepts that bargain before the window for a pragmatic framework closes.
Recommended!
To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.
Lawrence Solum