Algorithmic disclosure rules

Abstract During the past decade, a small but rapidly growing number of Law&Tech scholars have been applying algorithmic methods in their legal research. This Article does it too, for sake saving disclosure regulation failure: normative strategy that has long considered dead by scholars, conspicuously abused rule-makers. Existing proposals to revive duties, however, either focus on industry policies (e.g. seeking reduce consumers’ costs reading) or rulemaking simplifying linguistic intricacies). But failure may well depend both. Therefore, this develops `comprehensive approach', suggesting use computational tools cope with and behavioral failures at both enactment implementation phases thus filling void Law & Tech scholarship. Specifically, outlines how can be used holistic manner address many disclosures from parliament consumer screens. It suggests multi-layered design where lawmakers deploy three order produce optimal rules: machine learning, natural language processing, experimentation through regulatory sandboxes. To clarify why these tasks should performed, contexts online contract terms privacy are taken as examples. Because is frequently met well-justified skepticism, problems its compatibility legitimacy, efficacy proportionality also discussed.