Uri Benoliel is Professor of Law at the College of Law & Business and author of Disclosure Laws in the Digital Era: The Case of Franchising (Cambridge University Press, 2026).
In More Than You Wanted to Know: The Failure of Mandated Disclosure (Princeton University Press, 2014), Omri Ben-Shahar and Carl E. Schneider delivered a powerful critique of disclosure laws, one of the most widely used instruments of modern regulation. Across consumer finance, healthcare, and other fields, these laws require businesses and professionals to provide information so that individuals can make informed choices. Yet recipients often fail to read the disclosures, struggle to understand them, or cannot translate what they learn into better decisions. Ben-Shahar and Schneider questioned whether mandated disclosure was worth preserving at all.
More than a decade later, their critique remains highly relevant. Yet technological developments raise an important question: Can artificial intelligence and big data help overcome some of the longstanding limitations of mandatory disclosure?
That question motivates my new Cambridge University Press book, Disclosure Laws in the Digital Era: The Case of Franchising. The book argues that artificial intelligence and big data provide an opportunity to rethink both how we evaluate disclosure quality and how lawmakers should design disclosure requirements. Its central proposal, which I call Tech Disclosure, is to make legally mandated information accessible to computational analysis while using AI to improve the language in which that information is communicated.
Rethinking Disclosure Quality
Disclosure laws serve purposes extending beyond the transmission of information. By addressing information asymmetries, they seek to improve individual decisions, encourage competition, protect autonomy, and offer an alternative to more intrusive substantive regulation. Their effectiveness, however, depends on the quality of the disclosures they generate.
Scholars and regulators have traditionally assessed that quality through familiar criteria: document length, the cumulative number of notices, conventional readability formulas, and the prevalence of legal jargon. These indicators identify important weaknesses, but they do not fully account for the capabilities of contemporary technologies.
A lengthy disclosure may be impossible for a consumer to analyze independently, yet useful if its terms can be compared computationally with those of thousands of similar products. Conversely, a document that is easily downloaded may still be difficult to analyze accurately and efficiently using automated tools, particularly when its contents are scanned, inconsistently labeled, and poorly structured.
The book therefore proposes evaluating disclosure quality not solely through the burdens imposed on human readers, but also through the capacity of modern technologies to transform disclosed information into usable knowledge.
Big Data Readiness: Making Disclosures Accessible to Big Data Aggregators
The first pillar of Tech Disclosure proposes a fundamental shift in the intended audience of mandatory disclosure. Traditionally, disclosure laws have been designed primarily around individual consumers and investors, who are expected to receive, understand, and act upon disclosed information. This approach, however, overlooks the increasingly important role of big data aggregators, which possess technological capabilities that individual recipients generally lack.
Big data aggregators can collect and analyze information across thousands of disclosures, using advanced analytical tools, including AI, to identify patterns and anomalies, compare contractual terms and market offerings, and transform complex datasets into accessible, actionable insights. Through these functions, they can help consumers and investors make informed decisions without requiring them to examine extensive disclosure documents individually.
Accordingly, the book argues that disclosure regulation should be designed not merely to facilitate individual access to information, but also to accommodate the technological and operational needs of big data aggregators.
This requires two complementary reforms.
The first concerns disclosure database accessibility. Public disclosure databases should provide the infrastructure necessary for large-scale data collection, including standardized metadata, advanced search functionalities, bulk downloading, application programming interfaces (APIs), historical archives, and clear data-reuse permissions. These features would enable big data aggregators to collect and analyze disclosure information systematically rather than retrieve documents individually.
The second concerns disclosure machine readability. Disclosures should be presented in formats that permit reliable automated extraction and processing. This requires standardized document structures, consistent taxonomies, and unique identifiers for comparable information. Merely publishing disclosure documents online does not ensure that their contents can be effectively analyzed by computational systems.
The underlying principle is that the ultimate beneficiaries of disclosure law remain consumers and investors, but big data aggregators should become a central consideration in the design of disclosure requirements. By enabling sophisticated technological actors to process information at scale, disclosure law can reduce its traditional dependence on the limited cognitive capacities of individual recipients and better realize its objectives of informed decision-making, competition, and market transparency.
AI-Oriented Linguistic Standards: Improving Disclosure Language
The second pillar addresses the linguistic quality of individual disclosures.
Traditional readability measures, often based on word and sentence length, cannot adequately capture many deficiencies that make legal texts difficult to understand, including ambiguity, unnecessarily complicated syntax, redundancy, and unfamiliar terminology.
AI-assisted methods offer more sophisticated evaluation techniques. AI-powered writing assistants can identify grammatical and clarity problems; AI-based paraphrasing tools can suggest more concise formulations; and AI chatbots can identify specialized legal terminology that may impede comprehension.
These tools also enable more systematic and scalable assessments of disclosure language than conventional methods alone.
The book proposes translating these capabilities into regulatory practice through voluntary, rebuttable presumptions of compliance with statutory linguistic standards.
For example, a regulated entity could submit a favorable assessment from a qualifying AI writing assistant as initial evidence that its disclosure satisfies specified readability requirements. Comparable assessments could support presumptions concerning conciseness or the avoidance of excessive legal jargon.
Such presumptions must remain rebuttable. Linguistic simplicity does not necessarily establish legal accuracy. AI systems may overlook important qualifications, misinterpret technical expressions, or suggest revisions that alter substantive rights and obligations. Human review and independent regulatory scrutiny therefore remain essential.
The objective is not to replace legal expertise with automated judgment, but to provide more rigorous methods for evaluating and improving disclosure language.
Franchising as an Empirical Case Study
The book applies the Tech Disclosure framework to regulation under the Federal Trade Commission’s Franchise Rule. Franchising provides a useful setting because prospective franchisees receive extensive precontractual information before undertaking substantial financial commitments.
The empirical analysis examines government-related disclosure databases, the machine readability of disclosure documents, and the linguistic characteristics of selected provisions.
The findings reveal significant limitations in database accessibility, obstacles to automated information extraction, and substantial opportunities for AI-assisted improvements in readability, conciseness, and legal jargon.
These findings illustrate a broader structural problem: disclosure regimes may satisfy conventional legal requirements while remaining poorly adapted to contemporary technological capabilities.
Franchising thus serves as a case study for a regulatory framework with potentially broader applications to consumer finance, healthcare, securities regulation, insurance, and privacy.
Toward a New Architecture of Disclosure Law
Tech Disclosure does not assume that technology can overcome every limitation of mandatory disclosure. Consumers may disregard even clearly presented information, automated analyses may produce misleading results, and some contractual practices warrant substantive prohibition regardless of disclosure quality.
Nor does the framework necessarily displace alternative reforms involving simplified disclosure scores and information intermediaries (Oren Bar-Gill), or personalized disclosures (Ariel Porat and Lior Jacob Strahilevitz).
Its distinctive contribution is to shift attention from disclosure as a document-delivery obligation toward disclosure as an information infrastructure capable of supporting both computational analysis and individual comprehension.
Under this approach, lawmakers should consider not only what businesses must disclose and when, but also how the resulting information must be stored, structured, accessed, processed, and evaluated.
This shift has implications for contract law and consumer protection. It recognizes that the benefits of disclosure need not depend exclusively on individuals personally reviewing every document they receive. Properly designed disclosure systems can also support market-wide analysis, meaningful comparisons, and more accessible explanations of contractual terms.
The central question is therefore not simply whether individuals will read the information that businesses are legally required to provide. It is whether disclosure laws can be designed to ensure that such information becomes genuinely useful in an increasingly digital economy.
Ben-Shahar and Schneider demonstrated why traditional disclosure frequently fails. My book suggests that advances in AI and big data provide an opportunity to reconsider what disclosure regulation can accomplish.
The future of disclosure law may depend less on persuading individuals to read more information and more on designing regulatory systems that enable modern technologies to make that information genuinely useful.
Recent Comments