Chung on Structural Defect Framework

Jon Chung has posted to SSRN Structural Defect Framework. The abstract provides:

This Article introduces the Structural Defect Framework (SDF) to address a critical failure in modern liability doctrine: the law’s tendency to misclassify harms produced by invisible structural defects—architectural design choices hidden from end-users but fully controlled by institutional designers. Across four disparate case studies—adaptive streaming (Netflix-style bitrate attenuation), model quantization in legal AI, the systemic hollowing of Arizona’s victims’ “right to confer,” and unstructured discretion in Title VII employment decisions —the Article demonstrates how designers create loosely regulated spaces where critical parameters are under-specified yet entirely controllable. Within these spaces, fidelity loss in pixels, legal reasoning, and victim voice produces predictable, patterned harms that doctrine wrongly dismisses as mere optimization, random error, or individual malice.

The SDF proceeds in three layers. It begins with the threshold premise: that fidelity is a choice—a controllable design parameter, or “knob,” that exists and is turned. This premise is immediately followed by a tripartite gateway that engineers Structural Invisibility: (1) a fixed infrastructure or interface that stabilizes the user’s view while internal configurations change; (2) nominal equivalence, where distinct configurations are sold under a single label; and (3) pipeline opacity, where multiple intermediaries can attenuate fidelity without creating any audit point. Where these preconditions hold, a four-element diagnostic predicts where harm will occur and persist: (1) Structural Invisibility (the operator cannot audit the parameter); (2) Foreseeable Harm (the harm flows from a controllable configuration choice); (3) Doctrinal Mismatch (law immunizes the architecture from scrutiny); and (4) Epistemic Asymmetry (evidence of fidelity risk is held by the designer alone). By formalizing these patterns, the Article reframes AI safety, civil rights, consumer protection, and victims’ rights as parallel instances of designed-in harm.