TL;DR:
- The Advisory Committee on Evidence Rules held its Fall 2026 enrollment planning around AI generated evidence on October 15, 2026 in Boston. It withdrew the public-comment version of Rule 707 and advanced a revised Version 2 for further study. The Committee also convened panels on deepfakes (Rule 901(c)) and AI-evidence (Rule 707) with the Fall 2026 mini-conference identified as the next formal step. Practitioners should expect continued gatekeeping changes, potential pretrial notice requirements for AI-generated evidence, and evolving standards for reliability and authentication of machine-produced outputs. Objection Academy’s AI-evidence practice modules align with these developing standards to sharpen objections and trial-readiness in this space. Sources: U.S. Courts materials from October 2026, including the official evidence rules agenda book and supporting memoranda.
Trial lawyers preparing for federal litigation should take note of a concrete, recent shift in how AI-generated evidence could be treated at trial. On October 15, 2026, the Advisory Committee on Evidence Rules published its Fall 2026 agenda and related materials, signaling a continued evolution of rules addressing deepfakes and AI-produced material. The core developments are twofold: first, a revised approach to Rule 707 (the proposed rule governing AI-generated evidence offered without a human expert); second, a framework for authenticating deepfakes under a new Rule 901(c). The committee explicitly stated that it had withdrawn the prior public-comment version of Rule 707 and would circulate a revised Version Two for further study, with the expectation that Version Two would be the focus at the Fall 2026 mini-conference. In other words, the federal rules process is actively retooling the gatekeeping mechanism for AI produced evidence, and the next concrete milestones are set for late 2026. The minutes also confirm a panel discussion on deepfakes and AI-generated evidence as part of the October 2026 meeting and that the Fall mini-conference would bring Version 2 of Rule 707 and the 901(c) deepfake framework back for further public input. Practicing litigators should monitor these developments closely, because outcomes could affect pretrial disclosures, objections, and the admissibility of AI-derived conclusions or outputs, including those used for summaries, transcriptions, or document analysis. See the official October 2026 evidence rules materials for the precise scope and sequencing. (uscourts.gov)
What happened on the ground this fall
The October 15, 2026 agenda book confirms that the Advisory Committee has been actively wrestling with deepfakes and AI-generated evidence. It notes that a new Rule 901(c) would require an opponent to present enough information to support a finding that an item was fabricated by generative AI, after which the proponent would have to prove authenticity by a preponderance of the evidence. The panel discussions and memoranda accompanying the agenda reflect a phase of expert input to shape the draft rule text and notes. More importantly, the minutes reveal a concrete procedural move: at its last meeting, the Committee voted to withdraw the previously issued public-comment version of Rule 707 and to circulate a revised Version Two for further study. The Fall 2026 mini-conference would be the forum to take up Version Two and seek further input, suggesting a staged, not a single-shot, rulemaking process. The committee’s plan to address Rule 707 alongside Rule 901(c) at the Fall mini-conference underscores the intertwined nature of reliability standards and authenticity challenges in AI-produced evidence. The committee explicitly stated that discussions would continue at the Fall 2026 mini-conference with a focus on Version 2 of revised Rule 707, signaling a structured path toward potential publication and public comment in a subsequent cycle. (uscourts.gov)
Practical implications for working litigators
- Gatekeeping moves up a notch for AI evidence. The two-step approach under Rule 901(c) would require an opponent to raise a prima facie challenge to deepfakery before the court imposes a heightened authenticity inquiry. If the opponent meets that burden, the proponent must then show the item is authentic by more likely than not. This creates a more rigorous gatekeeping framework than traditional 901-based authentication alone and could affect what juries see without a live expert or Daubert-type showing. Practitioners should be prepared to articulate and separately verify the reliability or non-reliability of AI outputs, including whether an independent validation or reproducibility demonstration is available. The agenda notes emphasize that the standard is intended to address the increasing sophistication and availability of deepfake and AI-produced content. (uscourts.gov)
- Rule 707 revisions will shape when and how AI-generated conclusions can be admitted without an accompanying expert. Version Two of the revised Rule 707 is designed to balance accountability and practicality in civil and criminal settings, including considerations such as notices, potential carve-outs, and disclosure requirements. The Fall mini-conference will gather additional feedback, and practitioners should anticipate briefing and motion practice to reflect the evolving standard of reliability for AI-produced material. The minutes reveal that the committee is weighing how to harmonize the Rule 707 text with possible references to Rule 702 or to a broader, cross-cutting reliability framework, which may affect how a trial judge evaluates AI-generated outputs in real time. (uscourts.gov)
- Pretrial notice and credibility challenges gain prominence. The working drafts and panel discussions include issues around notice and the burden of going forward, meaning counsel should consider whether pretrial disclosures or notices are advisable whenever AI tools are used to produce material for the case. This aligns with the Committee’s emphasis on tailoring an approach that prevents last-minute authenticity fights and preserves opportunities to cure or challenge AI-generated material before trial. While not yet law, these signals suggest a shift in trial preparation habits to include early assessment of AI-generated outputs as potential evidentiary challenges. (uscourts.gov)
What trial teams should do now
- Track the Fall 2026 mini-conference outcomes. The Committee’s plan to take up Version Two of Rule 707 and the 901(c) deepfake framework at the Fall 2026 mini-conference means there will likely be a public-comment cycle and possibly a published draft or reporter’s memo. Prepare to adjust pretrial practice, objection strategies, and motion templates as soon as a final version is released. The official materials indicate that final text and notes could emerge after the Fall conference, shaping how AI evidence is addressed in 2027 trials. (uscourts.gov)
- Build a practical playbook for objections to AI outputs. Given the two-tier approach described in Rule 901(c), trial teams should practice: (a) early identification of potential deepfakes, (b) preserving objections for authenticating AI outputs, and (c) presenting a credible authenticating path if the opponent raises a deepfake challenge. Objection Academy offers objection drills and trial-readiness training about handling complex evidence issues, including AI-related questions, which can help trial teams prepare for evolving rules and argument structures as these proposals mature. See Objection Academy resources discussing AI-evidence considerations and practical objection strategies. (objectionacademy.com)
- Align discovery and trial-ready practices with evolving standards. If Rule 707 Version Two introduces notice requirements or special disclosures for AI-produced material, counsel will want to implement pretrial checklists and standardized disclosure language to cover AI-generated conclusions, outputs, and any supporting methodology. The Committee’s ongoing work, including the explicit linkage to deepfake authentication under Rule 901(c), points to a broader need for a coherent trial-readiness workflow around AI evidence. The October 2026 materials outline this trajectory clearly. (uscourts.gov)
Evergreen angle for ongoing trial-readiness
Even as the federal rules process continues, trial teams can stay ahead by training for AI-related evidence issues now. Objection Academy’s curriculum emphasizes real-world objection construction, evidence-applicability drills, and courtroom simulations that reflect current AI-evidence concerns. By integrating those exercises into trial teams’ practice routines, litigators can build a robust capacity to handle AI-evidence challenges, preserve objectivity, and maintain control over authenticating processes when the rules finally crystallize. The evolving landscape in 2026 confirms that AI-generated evidence is not a temporary concern but a persistent feature of modern litigation, making disciplined objection practice and evidence analysis essential components of trial readiness. (objectionacademy.com)
Sources
- Advisory Committee on Evidence Rules, October 15, 2026 agenda book and accompanying memorandum materials (Boston, MA). Key passages confirm withdrawal of the public-comment Rule 707 version and the plan to consider Version 2 at the Fall 2026 mini-conference, along with deepfake Rule 901(c) discussions. (uscourts.gov)
- Additional context and summaries from related public-facing materials and industry coverage referencing the October 2026 discussions and the anticipated Fall mini-conference trajectory. (uscourts.gov)
- Objection Academy coverage and training materials on AI-evidence and objection practice for trial attorneys. (objectionacademy.com)