How The MAD Act Digitizes The Federal Record
Title V: $1 Billion for Mass Digitization, Public Deposit, & AI-Ready Government Records
The National Archives holds 13.5 billion pages of textual records. As of today, approximately 3.4 percent have been digitized. The Library of Congress holds 178 million items across 470 languages — the largest library collection in the world. Most of the Government Publishing Office's digital collections begin with the 103d Congress (1993); the substantial historical record before that remains analog. Meanwhile, records on microfilm, nitrate film, thermofax paper, and acid-degraded paper are physically deteriorating every day. Title V of the MAD Act appropriates $1 billion to the National Archives, $30 million to the Library of Congress, and $20 million to GPO for mass digitization through competitively awarded contracts — with a mandatory requirement that contractors deposit complete copies with independent, nongovernmental public access platforms, ensuring permanent public access that doesn't depend on any government system surviving the next administration. The bill also modernizes records schedules to preserve records that have value for AI training and governmental automation — records that current appraisal criteria would authorize for destruction.
- The government can surveil citizens but citizens can't computationally analyze the government: Over successive decades, the federal government has significantly expanded its capabilities in intelligence collection, surveillance, data analysis, and information processing — while the vast majority of the government's own records remain inaccessible to the public in analog form. Congress found this asymmetry undermines the foundational principle that an informed public is the ultimate check on governmental power.
- AI will deepen or correct this asymmetry — depending on who has the data: If government records remain analog, the benefits of AI will accrue disproportionately to those with institutional access — the government itself, well-resourced private entities, and foreign adversaries — while the public's capacity to oversee its own government will continue to erode. Digitization and public release would enable citizens, researchers, journalists, and civil society to apply the same analytical capabilities to the work of government.
- Physical deterioration is irreversible: Microfilm, nitrate-based film, thermofax paper, acid-degraded paper — materials with a projected usable lifespan of fewer than 50 years are deteriorating now. Absent timely digitization, substantial volumes of historically significant material face permanent loss.
- The sunk cost argument: These records represent hundreds of years of institutional creation, archival appraisal, and public investment in long-term preservation — individually evaluated and selected for permanent retention because of their enduring value. The additional investment to digitize them is modest relative to the costs already borne in their creation and preservation. Digitization would for the first time unlock their full value.
- Scale comparison — Google Books: The largest comparable mass digitization effort in history required 15 years and an estimated $750 million to scan approximately 40 million volumes — a fraction of the federal government's analog holdings.
- Three covered institutions, coordinated: NARA (all analog holdings), GPO (historical publications predating the 103d Congress), and the Library of Congress (at-risk materials, government documents, congressional records, legal materials, Law Library holdings). Within one year of enactment, the three institutions must execute a memorandum of understanding establishing joint planning, common procedures, and shared reporting. No contracts may be awarded until the MOU is signed.
- Full and open competition: All contracts for digitization services are awarded through competitive procurement under the Federal Acquisition Regulation. No exclusive arrangements. The government retains unlimited rights in all digitized records — no contractor gets exclusive commercial access to public domain material.
- Career Digitization Program Directors: Each institution gets a dedicated career director removable only for cause, with direct authority over contract awards, expenditures, and quality assurance. NARA's Director of Federal Records Digitization is a career Senior Executive Service position requiring at least 7 years of experience in archival science, digital preservation, or large-scale digitization. The LOC and GPO directors each require at least 5 years of relevant professional experience and can be removed only for cause, with 30 days' advance notice to Congress. None are political appointees.
- Prioritization order: (1) At-risk records on deteriorating media. (2) Declassified records from the National Declassification Center not yet available online. (3) High-demand records based on reference request and FOIA data. (4) Administrative and program records with AI/automation potential. (5) Historical government publications predating the 103d Congress. (6) Everything else.
- Milestone targets: At-risk records condition survey within 2 years. All at-risk records digitized within 10 years. All processed declassified records digitized within 12 years. All pre-103d Congress GPO publications digitized within 15 years. Annual reporting on progress through fiscal year 2051.
- Every contract requires public deposit: Every digitization contract must include a requirement that the contractor, at its own expense, deposit a complete, unaltered copy of all production masters and metadata with at least one independent public access platform — a nongovernmental repository providing free, unrestricted public access. Deposit must occur within 365 days of completing digitization.
- Public access preference in procurement scoring: Proposals from contractors who themselves operate (or are affiliated with) an independent public access platform receive significant additional weight as a scored evaluation criterion. Contractors who commit to maintaining a complete copy on their platform for the duration of its operation satisfy the deposit requirement automatically.
- Open formats, no restrictions: All deposited materials must be in open, machine-readable formats — available to the public free of charge, without restriction on access, download, or reuse. Production masters in uncompressed or losslessly compressed TIFF or equivalent archival format. Access copies in PDF/A, JPEG 2000, or similar open formats.
- No exclusive rights, no embargoes: No contract may confer exclusive rights on any contractor. The government retains unlimited rights in all digitized records. No embargo or period of exclusive commercial access may be imposed on public domain records digitized after enactment. Existing partnership agreements are grandfathered, but no new exclusive arrangements may be created.
- The problem: Current records schedules — the rules governing what gets kept and what gets destroyed — don't account for the prospective value of records as training data, reference corpora, or knowledge bases for AI systems. Records documenting how agencies make decisions, interpret statutes, and carry out their missions are routinely authorized for destruction before their computational value can be realized.
- What the bill requires: The Archivist must update appraisal guidelines to require preservation of records documenting: administrative and operational decisions in the execution of statutory mandates; the reasoning and deliberative processes underlying adjudicative determinations and policy interpretations; standard operating procedures and institutional workflow documentation; and any additional categories with significant value for automation, augmentation, or computational analysis of government functions.
- Comprehensive review within 3 years: The Archivist must review all existing General Records Schedules and agency-specific schedules — in consultation with OMB and OSTP — to identify records currently authorized for destruction that should be preserved under the new criteria. Revised schedules must be issued. Records already authorized for destruction but not yet destroyed can be saved.
- Deliberative process protected: Preservation of decision-making records doesn't require attributing contributions to individual employees. Agencies may redact individual identifying information. Nothing in this section waives FOIA deliberative process exemptions or requires public disclosure of exempt records.
- Facility protection: No facility used for the storage, staging, or digitization of records under this title may be sold, transferred, repurposed, or closed during active contracts — unless the institution head provides 180 days' advance written notice to Congress with an impact assessment and relocation plan. Congress can block the action with a joint resolution of disapproval under expedited procedures.
- Deposit enforcement: Every contract includes a material breach clause for failure to complete public deposit on time, a liquidated damages schedule for delayed public access, and a provision allowing the covered institution to complete the deposit itself using the contractor's production masters — and recover the costs from the contractor's payments.
- Supplement not supplant: The appropriation is in addition to existing budgets. No covered institution may reduce its budget request for existing activities because digitization funding has been made available. Funds may be used exclusively for digitization activities — not for general administration, facilities maintenance, or existing programs.
- IG audits and major contract review: Biennial Inspector General audits of all expenditures. Contracts exceeding $10 million receive mandatory IG review within 60 days of award for compliance with competitive procurement, public access requirements, and use-of-funds restrictions. Results reported to Congress and made public.
- Whistleblower protection: Employees and contractors who disclose waste, fraud, abuse, mismanagement, or noncompliance with this title are protected from retaliation under existing federal whistleblower statutes.
- Sustained retention planning: By September 30, 2049, the Archivist must submit a plan for funding digital storage and retention of all digitized materials beyond fiscal year 2051 — including annual cost estimates, infrastructure requirements, and legislative recommendations.