TL;DR:
The Advisory Committee on Evidence Rules issued a May 17, 2026 memorandum stating it would not advance the proposed Federal Rule of Evidence 707 at this time. The Committee revised the proposed rule and will study AI related issues, including deepfakes, at its next meeting. This keeps AI-generated outputs within the existing evidentiary framework for now, while signaling ongoing scrutiny of machine-generated evidence. For trial teams, the practical effect is a continued need to ground AI-derived materials in traditional Rule 702/901 foundations and to prepare robust objections and foundations until a formal rule is adopted. Public comments had closed on February 16, 2026, and the next steps will unfold in the Committee’s June 2026 discussions. (uscourts.gov)
What happened
On May 17, 2026 the Advisory Committee on Evidence Rules released a memorandum detailing its May 7 2026 meeting. The core takeaway: the Committee does not recommend action on the proposed Rule 707 governing AI-generated evidence offered without an expert witness. Instead, the Committee indicated it had revised the proposed Rule and planned to continue study of AI related issues, including “the problems posed by deepfakes,” at its next meeting. This formal decision effectively pauses the immediate adoption process for FRE 707 and preserves the status quo while signaling a continued policy debate. The memorandum corroborates that Rule 707 was recognized and discussed earlier in 2026, including a public-comment window that closed on February 16, 2026, and that the Committee intends to revisit the topic. (uscourts.gov)
Context for practitioners: the proposed FRE 707 anticipated a rule tailored to machine-generated outputs, aiming to treat AI-produced material with the same reliability considerations as other expert-like evidence when offered without a human expert witness. The public dialogue included industry commentary and formal comments, highlighting concerns over breadth, practicality, and the potential scope of AI evidence in routine litigation. While supporters argued for a principled Daubert-like gate, opponents warned of overreach into ordinary data and routine digital outputs. (ixsor.com)
Why it matters to trial attorneys
The stall on FRE 707 keeps courts’ screening of AI-derived material squarely in the traditional Rule 702 context for now. If FRE 707 had advanced as proposed, courts would have faced a new gatekeeping framework specifically for machine-generated outputs lacking a human expert. By delaying action, litigators should rely on established standards for authentication, reliability, and admissibility of AI-derived content, including:
- Foundation and authentication under Federal Rules of Evidence 901 and 902, where applicable, to establish provenance and integrity of AI outputs.
- Reliability and relevance analysis under Rule 702, including Daubert-style considerations if AI output is presented as opinion or inference without a live expert.
- Caution around “deepfakes” and other AI-manipulated media, which the Committee plans to study further, recognizing the heightened risk of misattribution, misrepresentation, or software-driven bias. (uscourts.gov)
For trial teams, the takeaway is pragmatic: continue to treat AI-generated content as evidence subject to the same foundational requirements as other digital evidence. Expect continued need for careful disclosure, source-tracking, and, where appropriate, testimony from an independent expert about data provenance, training sources, and the limitations of the AI tool. Public commentary on the topic, including critiques from advocacy groups and industry groups, underscores the importance of concrete, verifiable foundations when AI outputs are introduced at trial. (justice.org)
Bloomberg Law’s coverage framed the decision as a stall that may be viewed as a prudent delay rather than a setback, noting the legal community’s ongoing debate about whether new gatekeeping is needed for AI outputs. That framing aligns with the Committee’s approach: pause, revise, and study before formal adoption. Practitioners should monitor the next ACER deliberations for any revised text or new guidance about AI evidence and deepfake authentication. (news.bloomberglaw.com)
Practical steps for trial teams now
- Preserve AI artifacts and metadata. Maintain a robust, auditable trail of how an AI output was generated, including model name, version, prompts, and any post-processing steps. This supports authentication and cross-examination if the evidence is challenged.
- Build a rigorous foundation for AI outputs. Even absent FRE 707, ensure AI-derived material can be treated as non-expert evidence under Rule 702 where appropriate, with a clear link to data sources, input data, and the method of generation.
- Prepare for Daubert-style challenges. If an opposing counsel presents AI outputs as authoritative conclusions, be ready to introduce a qualified expert who can explain limitations, error rates, and potential biases in the tool.
- Plan objections strategically. Practice timely, precise objections to AI-generated content and its foundation, including challenges to authenticity, reliability, and relevance. Objection Academy can be a practical training resource to simulate objections to AI-derived materials and to rehearse courtroom responses in a risk-free setting.
- Stay ahead on “deepfake” issues. Given the Committee’s stated interest in deepfakes, counsel should develop a playbook for authenticating manipulated media and for arguing real-time reliability concerns in voir dire, opening, or during trial.
- Consider the role of expert testimony. If AI outputs are integral to the case, evaluate whether a live expert is needed to interpret outputs for the jury, or whether the outputs can be admitted under an expert-witness framework with appropriate disclosures.
These steps align with the current evidentiary landscape and position trial teams to respond quickly as FRE 707 guidance evolves. Objection Academy can play a constructive role in building muscle around objecting to AI-derived evidence and in rehearsing courtroom scenarios where AI artifacts are at issue.
What to watch next
The Advisory Committee intends to continue study at its next meeting, with particular attention to AI issues including deepfakes. While no specific date is published in the May 2026 memorandum, the committee’s public-facing materials indicate ongoing work through mid-2026. Law firm analyses and coverage suggest the rulemaking process may resume with revised text or additional commentary before any final vote. For trial teams, this means ongoing monitoring of ACER’s public materials and any public comment opportunities tied to FRE 707 or related AI-evidence topics. (uscourts.gov)
Objection Academy angle
In a year where AI-generated materials increasingly surface in trials, Objection Academy remains a practical resource for trial teams preparing to handle objections to evidence grounded in machine-generated outputs. Training modules focused on identifying reliable foundations, crafting precise objections, and simulating courtroom objections to AI evidence can help litigators maintain trial-readiness even as rules adapt to new technologies. As FRE 707 undergoes further study, integrating targeted objection practice around AI-derived materials ensures courtroom performance remains sharp when the rules finally crystallize.
Sources
- Advisory Committee on Evidence Rules May 2026 memorandum, May 17, 2026 (report of the May 7, 2026 meeting) confirming no action on Rule 707 and plans for further study, including deepfakes; next meeting to consider AI issues. (uscourts.gov)
- Advisory Committee on Evidence Rules May 2026: agenda/books and related materials (May 7 2026 meeting context). (uscourts.gov)
- Public-comment timeline and context for FRE 707, including February 16, 2026 deadline for public comments. (justice.org)
- Bloomberg Law analysis of the stall and ongoing debate around FRE 707. (news.bloomberglaw.com)
- Primary background on what Rule 707 would have addressed, and the AI-evidence framework proposed by the Advisory Committee. (ixsor.com)
Note: The article above is based on verifiable, publicly available materials from the Judicial Conference and reputable legal coverage. It focuses on a concrete, timely development with actionable implications for trial teams, while also highlighting practical training considerations for handling AI-related evidence in court.