The Behavioral Premise: Why AI Adoption Failures are Misdiagnosed

Abstract

This paper asks why so many AI initiatives that stall, or never earn sustained use, are blamed on the wrong causes. It describes a diagnosis reflex: when adoption disappoints, organizations look first to the technology, then to the organization around it, then to governance, and only rarely to the behavior of the people expected to rely on the system. The paper tests those four explanations against the evidence and sets out the Behavioral Premise, a formal statement that adoption outcomes are determined in large part by how people actually decide whether, when, and how much to rely on AI, with theoretical foundations drawn from behavioral science. Because a premise is only useful if it can fail, the paper states falsifiable predictions along with the conditions under which it would be wrong, addresses the strongest objections, and offers notes on how the argument applies in different sectors. It closes by describing what changes in practice when the premise is accepted, illustrated through worked examples.