TL;DR
The Advisory Committee on Evidence Rules declined to advance proposed Rule 707 on May 7, 2026. The December 2027 effective date that attorneys have been citing is no longer realistic, and the rule may return for further revision before it moves anywhere. Louisiana’s Act 250, in force since August 1, 2025, already requires attorneys to verify evidence’s authenticity before offering it to a court. New York’s Part 161 and California’s Rule 10.430 also touch AI in the courtroom, but each one regulates something different. Attorneys need to know what each rule covers before assuming any of them apply to their evidence.
The Advisory Committee on the Federal Evidence Rules met on May 7, 2026, and chose not to advance proposed Rule 707 to the Standing Committee. This is the required next step before any rule change reaches the Judicial Conference, the Supreme Court, and Congress. The proposal had drawn more than seventy written comments and oral testimony in January. The Committee’s agenda described its options as revision, another comment period, or further study. Advancement was not one of them.
Proposed Rule 901(c) fared no better. The Committee declined to publish it for comment at all, after studying it across six meetings. December 2027 was already the earliest possible outcome under a smooth path through the rulemaking process. That path is gone for now. Attorneys tracking a federal timeline are tracking a date that no longer applies.
What Do Some State Rules Cover?
While Louisiana, New York, and California all regulate different aspects of AI, only Louisiana’s rules directly address the admissibility and verification of evidence.
Louisiana’s Act 250 imposes an actual verification duty on attorneys. The law took effect on August 1, 2025, amending the Louisiana Code of Civil Procedure, Article 371, to require attorneys to exercise reasonable diligence in verifying the authenticity of evidence before offering it to a court.
An attorney who knew, or should have known through reasonable diligence, that evidence was false or manipulated faces contempt findings and possible discipline for offering it without disclosure. The law includes a demonstrative exhibit exception, so AI-assisted visual aids used to illustrate testimony sit outside the duty that applies to substantive evidence. Act 250 does not change the standard for whether evidence is admissible. It adds a pretrial procedural duty that an attorney has to satisfy before an admissibility argument ever starts.
New York’s Part 161 regulates filings, not evidence. The rule took effect June 1, 2026, and requires attorneys who use AI tools to prepare papers submitted to a New York court to independently confirm that those filings contain no fabricated or fictitious cases, statutes, or other material. Sanctions and disciplinary exposure under existing frivolous-filing and candor rules back that requirement. The rule’s own text excludes materials offered as evidence, reserving those for separate treatment. Bills addressing AI evidence admissibility remain pending in committee in New York and have not been enacted.
California’s Rule 10.430 regulates the courts themselves. The rule took effect on September 1, 2025, and requires California courts that permit the use of generative AI by judicial officers and court staff to adopt a formal use policy addressing confidentiality, bias, and accuracy. Judges and court personnel are the audience for this rule. It sets no evidentiary standard for attorneys who bring AI-generated material into a case. An attorney practicing in California should know the rule exists, largely because it signals how the state’s courts are approaching AI more broadly. It will not decide whether a piece of AI-assisted evidence gets admitted.
An attorney handling a matter in Louisiana today operates under a real duty to verify evidence before offering it. An attorney handling a similar matter in a state with only a filing-accuracy rule or an internal court-governance policy is not covered by anything comparable. That attorney is simply untested in a set of rules that differ from one jurisdiction to the next rather than converging toward a single federal standard.
What Does a Defensible Verification Process Look Like?
A defensible verification process documents diligence as evidence comes in, rather than reconstructing it after a challenge forces the issue.
Chain of custody has to be built as the collection happens. Where a file came from, who touched it, and what happened between collection and production should be recorded in real time, since that record carries weight later and is nearly impossible to reconstruct convincingly after the fact.
Metadata has to be preserved deliberately. Native file metadata is often the clearest proof that nothing changed after a document or image was created, and it disappears permanently once stripped, whether through a careless file transfer or a well-meaning custodian trying to help.
A verification memo has to exist from day one. A record created the day evidence comes in reads differently to a judge than one assembled after opposing counsel raises a challenge, and a memo need not be long to carry weight. It needs a date that predates any dispute.
Client-submitted material needs the same scrutiny as anything else. Screenshots, recordings, and images that clients hand over voluntarily face the same authenticity questions as formally produced discovery, usually with far less documentation attached from the start.
A verification workflow has to match the actual rule in play. A process built for a Louisiana matter should not be assumed to satisfy a comparable standard in New York or California, since no comparable standard currently exists in either state. A multi-jurisdictional practice needs a process built to the strictest duty that could apply, not the most convenient reading of a rule that may not reach evidence at all.
Why Does Verification Matter Beyond the Courtroom?
A strong verification habit prevents problems long before a hearing date arrives. It keeps depositions on schedule instead of derailing them with an unplanned forensic detour. It keeps a case argued on its merits rather than giving opposing counsel an opening to attack the process. An attorney who has already demonstrated diligence walks into a hearing with an answer ready, rather than assembling one under pressure while a judge waits.
Clients benefit from this too. Many assume anything digital is automatically usable in court until a challenge proves otherwise in the middle of a case. An attorney who can explain what verification requires and which rule governs a given matter gives a client a clearer picture of the risk while there is still time to act.
How Can Complete Legal Help Build a Verification Workflow?
Complete Legal helps legal teams build collection and chain-of-custody practices that hold up when authenticity is questioned, documenting diligence at the point it occurs rather than reconstructing it later. With the federal rule stalled and state rules covering different ground in Louisiana, New York, and California, matching a workflow to the correct standard takes more than a general policy. If a matter touches any of these states, or another state moving in this direction, the conversation is worth having now, while there is time to build the process rather than defend a gap in one.
Reach out to Complete Legal to talk through what a verification workflow should look like for your practice.
Contact Us:
📧 info@completelegal.us | 📞 888-885-9909
Key Takeaways
- The Advisory Committee on Evidence Rules declined to advance proposed Rule 707 on May 7, 2026. The December 2027 effective date is no longer realistic, and further revision or comment may come first.
- Louisiana’s Act 250, effective as of August 1, 2025, is the only state rule that imposes a genuine pretrial duty to verify the authenticity of evidence, with a carve-out for demonstrative exhibits.
- Act 250 leaves the admissibility standard itself untouched. It adds a procedural duty on top of it.
- New York’s Part 161, effective June 1, 2026, governs the accuracy of AI-assisted court filings, not the authentication of AI-generated evidence. Bills on evidence admissibility remain pending in committee.
- California’s Rule 10.430 governs how judges and court staff use AI internally and sets no evidentiary standard for attorneys.
- An attorney needs to know precisely what a given state rule covers before assuming it applies to a specific piece of evidence.
