ABSTRACT
Picture a public-facing generative AI chatbot on a government website that can answer questions about benefits eligibility. Systems like these expose a sociotechnical-legal divide: they route around the standard due process framework, because there has been no formal government determination, yet they induce reliance from users, above and beyond past rounds of automated legal guidance. Moreover, if there is an error, current doctrine on government errors and equitable estoppel makes relief unlikely.
This Essay focuses on government deployment of public-facing generative AI chatbots and critically assesses the front-end power dynamics that determine who is filtered out of government programs before traditional procedural protections even attach. The only reliable backstop, right now, is what I call ‘clickwrap accountability’: a disclaimer, borrowed from contract law, that shifts responsibility for the tool’s performance to its users.
Importing this private law accountability measure into public law is a mistake. And we should know better, given the deficiencies of ‘notice and choice’ consent paradigms in information privacy law and mounting critiques of clickwrap contracts, in general. At a moment when more and more government actors are deploying generative AI chatbots for public-facing services, and when the federal government continues to push for greater AI adoption, we must insist on better public law protections than clickwrap accountability.
Solow-Niederman, Alicia, Clickwrap Accountability (June 15, 2026), 95 Fordham Law Review (forthcoming 2026); GWU Legal Studies Research Paper No 2026-45; GWU Law School Public Law Research Paper No 2026-45.
Leave a Reply