Executive Abstract
Purpose: This Model Local Rule addresses the verification crisis in family court by establishing standards for AI transparency and digital evidence provenance. It is designed for adoption by state and local courts as a local rule, general order, or standing order.
Core Components:
- Mandatory AI Disclosure - All parties filing documents must disclose AI use in research, citation generation, or drafting. Disclosure is made through a simple certification form attached to filed documents.
- Human-in-the-Loop Certification - Parties certify under penalty of perjury that all citations have been verified against authoritative legal databases and all factual claims have been independently confirmed. AI assistance is permitted; blind reliance is not.
- Evidence Passport Standards - Digital evidence submissions must include provenance documentation: capture timestamps, metadata preservation, chain of custody records, and disclosure of any editing or processing. Courts may require Evidence Passport documentation for contested exhibits.
- Sanctions Framework - Graduated sanctions for violations: warning and remedial education for first offense with good-faith error; monetary sanctions and referral to bar counsel for knowing fabrication; case sanctions (striking documents, adverse inference) for egregious or repeat violations. Safe harbor protects disclosed, verified AI use.
Implementation Roadmap: The rule contemplates four-phase adoption:
- Phase 1 (Months 1-3): Voluntary Disclosure Period - Courts encourage but do not mandate AI disclosure. Educational materials distributed to bar and self-help centers. No sanctions for non-disclosure during this phase.
- Phase 2 (Months 4-6): Mandatory Disclosure, Soft Enforcement - AI disclosure becomes required. Courts issue warnings for non-compliance but allow cure without sanction. Focus on education and building compliance habits.
- Phase 3 (Months 7-12): Full Enforcement with Good Faith Safe Harbor - Sanctions apply for fabrication and knowing misrepresentation. Good-faith errors with prompt correction receive warnings only. Evidence Passport standards applied to contested digital evidence.
- Phase 4 (Year 2+): System Integration - Verification tools integrated into court e-filing portals. Automated citation checking where feasible. Evidence Passport templates available through court self-help resources.
Intended Audience: Family court judges, court administrators, state judicial councils, bar associations developing AI ethics guidance, and self-represented litigant advocates.
Implementation Authority: This rule can be adopted as a local rule under most state rules of civil procedure, as a general administrative order, or as a standing order issued by presiding judges. It does not require legislative action in most jurisdictions.
Model Rule Text
LOCAL RULE [X]: VERIFICATION OF GENERATIVE AI USE AND DIGITAL EVIDENCE
Section 1. Disclosure of Generative AI Use
(a) Any party filing a document prepared with the assistance of generative artificial intelligence tools shall disclose such use by including the following certification:
(b) Disclosure is required regardless of the extent of AI use. Minor grammatical or formatting assistance, citation lookups, and substantive legal research all trigger the disclosure requirement.
(c) Self-represented litigants shall be provided with a simple disclosure form available at the Clerk of Court office and on the court's website.
Section 2. Verification Requirements
(a) All legal citations in filed documents must be verified by the submitting party against authoritative legal sources (e.g., official case reporters, Westlaw, LexisNexis, Google Scholar, or official court websites).
(b) Verification includes confirming: (1) the cited case exists; (2) the citation accurately reflects the holding; (3) the case has not been reversed or superseded; and (4) the legal standard cited is applicable in the relevant jurisdiction.
(c) Parties certify compliance by signing the disclosure certification or, if AI was not used, by signing the standard certification of compliance with court rules.
Section 3. Evidence Passport for Digital Evidence
(a) When digital evidence is contested, the court may require the submitting party to provide an Evidence Passport documenting:
- Capture method and timestamp
- Device and software used to create or capture the evidence
- Chain of custody from capture to court submission
- Any editing, processing, or AI involvement in creation
- Basis for authentication (personal knowledge, metadata, third-party verification)
(b) Evidence Passport documentation may be submitted in the form provided by the court or as a sworn declaration meeting the requirements of this section.
(c) Failure to provide required Evidence Passport documentation when ordered by the court may result in exclusion of the evidence.
Section 4. Sanctions
(a) Violations of this rule may result in:
- Written warning and requirement to complete AI verification training (first offense, good faith error)
- Monetary sanctions not to exceed $1,000 for knowing violations
- Striking of offending documents from the record
- Adverse inference instruction to finder of fact
- Referral to bar counsel for professional discipline (attorneys only)
- In egregious cases, case sanctions including dismissal or default judgment
(b) Safe Harbor: Parties who disclose AI use and certify verification in good faith are protected from sanctions if unintentional errors are later discovered, provided prompt correction is made upon notice.
(c) The court shall consider the party's resources, sophistication, and access to legal assistance when determining appropriate sanctions.
Implementation Guidance for Courts
Starting Point: Courts should begin with the voluntary disclosure period (Phase 1) to build awareness and provide educational resources before enforcing compliance requirements.
Educational Materials Needed:
- Sample AI disclosure certification forms
- Plain-language guidance on what counts as 'AI use'
- List of acceptable citation verification sources
- Evidence Passport template and instructions
- FAQ sheet addressing common questions from pro se litigants
Self-Help Center Integration: Court self-help centers should provide:
- Access to legal research databases for citation verification
- Training on using AI tools responsibly
- Evidence Passport preparation assistance
- Sample declarations and certifications
Bar Association Coordination: Work with local bar associations to:
- Develop CLE programs on AI verification compliance
- Create pro bono citation verification clinics
- Establish attorney mentor programs for pro se litigants
Policy Downloads and Resources
Interested in Piloting Verification Workflows?
ThreadLock works with courts, legal aid organizations, and bar associations to pilot structured evidence verification infrastructure. Contact us to discuss implementation in your jurisdiction.
Contact: policy@threadlock.ai
Citation and Attribution
Suggested Citation: Moore, Hannah. "Model Local Rule: Verification of Generative AI and Digital Evidence." ThreadLock Policy Research (2026).
License: This Model Local Rule is released under Creative Commons CC BY 4.0. Courts are free to adopt, modify, and distribute this rule without permission. Attribution appreciated but not required.
Speaking & Policy Engagement
Hannah Moore speaks on AI verification, digital evidence provenance, and judicial infrastructure modernization.
For speaking inquiries, panel participation, or policy consultation: contact@threadlock.ai
Frequently asked questions
QCan courts adopt this rule without legislative approval?▾
QDoes this rule prohibit AI use?▾
QHow does this affect self-represented litigants?▾
QWhat counts as 'AI use' requiring disclosure?▾
QWhat is the Evidence Passport requirement?▾
Sources
- Federal Rules of Civil Procedure — Foundation for local rule authority and procedural standards
- Federal Rules of Evidence - Rule 901 — Authentication standards for evidence
- Model Rules of Professional Conduct — Professional responsibility standards for attorneys
- Judicial Technology and AI Guidelines — Emerging standards for AI use in courts