Bodycams as PII/CUI

Abstract Police body-worn cameras are sold as tools for transparency, evidence, and public trust, yet current programs quietly place the recordings under the control of the very institutions they are supposed to constrain. This Article argues that the recurring problems around activation, retention, and access are not mere implementation glitches but the predictable result of misclassifying what body-worn camera footage is and who should control it. In practice, these recordings are dense bundles of personally identifiable information, intimate scenes of civilian life, and sensitive operational details, even as law and policy insist on treating them simultaneously as case evidence, public records, and repurposable media content. The Article develops the concept of “subjects of accountability” to describe officers and agencies whose conduct cameras are meant to document, and shows how they currently dominate every control point in the record: when and where cameras run, how long footage is kept, who may see it, and which clips are curated for public consumption. Drawing on federal guidance for protecting digital evidence and Controlled Unclassified Information, the Article argues that storage architecture is itself an ethical choice, and that the prevailing client-side model structurally favors quiet loss and selective disclosure. It proposes a server-side model with independent custody, cryptographic integrity checks, and auditable access logs, treating body-worn camera recordings as de facto warrants and scene logs whose disappearance should carry evidentiary presumptions and sanctions. The conclusion calls for reclassifying body-worn camera footage as sensitive evidence and personal data held in trust, prohibiting monetization of civilian images, tightening public-release rules, and conditioning camera deployment on independent technical and legal safeguards aligned with accountability rather than content production. New version and publication date for PDF reformat. Made in collaboration with ChatGPT 5.1 Thinking.