NYCLA Federal Courts Committee Comments on Proposed Federal Practice Rule Amendments

February 10, 2026

 

Via Electronic Submission
Honorable John D. Bates, Chair
Committee on Rules of Practice and Procedure

Re: Response to Request for Public Comment on Proposed Local Rules Amendments

Dear Judge Bates:

The New York County Lawyers Association’s Federal Courts Committee (the “Federal Courts Committee”) respectfully submits this letter in response to the Judicial Conference Committee on Rules of Practice and Procedure’s invitation to provide comments on the proposed amendments to existing rules and forms announced in a memorandum dated August 15, 2025 (the “Proposed Changes”). ¹ See Preliminary Draft, Proposed Amendments to the Federal Rules of Appellate, Bankruptcy, Civil, and Criminal Procedure, and the Federal Rules of Evidence (Aug. 15, 2025) (the “Proposed Changes”).²

The Federal Courts Committee sets forth below its comments to (1) the proposed changes to Federal Rule of Criminal Procedure 17; (2) the proposed Federal Rule of Evidence 707; and (3) the proposed changes to Federal Rules of Civil Procedure 26 and 45. The Federal Court Committee thanks the Committee on the Rules of Practice and Procedure for the opportunity to comment on the proposed amendments and the Rules Advisory Committees for their diligent work in preparing these amendments.

I. Proposed Changes to Federal Rule of Criminal Procedure 17

The Federal Courts Committee commends the Advisory Committee on Criminal Rules (the “Criminal Rules Committee”) on its decision to modernize Federal Rule of Criminal Procedure 17 (“Rule 17”). However, the Federal Courts Committee believes that the additional changes are needed beyond those reflected in the Report of the Advisory Committee on Criminal Rules dated May 15, 2025 (“Report”).

The Federal Courts Committee believes that there are numerous problems with the present rule, including that it has been inconsistently applied, overly restricts a criminal defendant’s ability to obtain discovery from third parties, and does not account for changes in the nature of discovery since its original adoption in 1944. We therefore endorse the Criminal Rules Committee’s decision to review and propose amendments to Rule 17.

In particular, the Federal Courts Committee notes its agreement with the Criminal Rules Committee’s decision to clarify that Rule 17(c) applies not only to trials but also permits parties to subpoena materials in connection with sentencing, suppression, detention, and revocation hearings. As the Criminal Rules Committee noted, parties often present evidence at these types of proceedings, and the Federal Courts Committee agrees that, to the extent there is a need for third-party evidence, the parties should have the opportunity to obtain it through a Rule 17(c) subpoena.

The Federal Courts Committee also agrees with the Criminal Rules Committee’s decision to modify the test that the Supreme Court articulated in United States v. Nixon, 418 U.S. 683 (1974), but believes that the Criminal Rules Committee’s proposal to “incremental[ly] change” the Nixon standard does not go far enough to remove the impediments to a criminal defendant’s ability to mount an effective defense and to rectify the unfair imbalance between the prosecution and defense in preparing for trial. In particular, as discussed below, the Federal Courts Committee believes that the Criminal Rules Committee should have adopted the proposal of the New York City Bar’s White-Collar Committee to replace the Nixon “admissibility” requirement with a standard that permitted both the prosecution and the defense to obtain Rule 17(c) discovery that is “relevant and material to the preparation of the prosecution or defense” or, at the very least, should have approved the “likely to lead to admissible evidence” standard supported by a minority of the Criminal Rules Committee. See Report § II.A.2.

As the Criminal Rules Committee noted, almost all federal courts now apply the Nixon test to restrict the scope of Rule 17(c) subpoenas. The Nixon test generally requires the subpoena proponent to “clear three hurdles,” specifically, (1) relevancy; (2) admissibility; and (3) specificity. Nixon, 418 U.S. at 700.

The Federal Courts Committee agrees with the Criminal Rules Committee’s decision to replace Nixon’s specificity requirement with a “reasonable particularity” standard. See Report § II.A.2. Multiple courts have recognized the unfairness of the current specificity requirement, which some courts have interpreted in a manner that conditions a criminal defendant’s ability to obtain needed documents that he/she has never seen before on the ability to adequately specify their contents. The Federal Courts Committee agrees with then-District Judge Richard J. Holwell’s observation that “requiring the defendant to specify precisely the documents he wants without knowing what they are borders on rendering Rule 17 a nullity.” United States v. Rajaratnam, 753 F. Supp. 2d 317, 321 n.1 (S.D.N.Y. 2011). The Federal Courts Committee therefore endorses the Criminal Rules Committee’s decision to replace the specificity requirement with a “reasonable particularity” standard, which mirrors the standard applicable to civil document discovery under Federal Rule of Civil Procedure 34(b).

As to the proposed modification of Nixon’s admissibility requirement, the Federal Courts Committee believes that the Criminal Rules Committee’s proposed change to a “likely admissible” standard does not go far enough to eliminate the problems with Nixon’s admissibility requirement. As the Criminal Rules Committee noted in its Report, courts have applied inconsistent standards to this element, with some courts applying a stricter standard requiring “the requesting party to prove with certainty that the information would be admitted.” See Report § II.A.2. This interpretation has often unfairly prevented defendants from obtaining exculpatory evidence, including impeachment material, leading practitioners to correctly opine that “courts have interpreted 17(c) so narrowly that it is rarely useful to criminal defendants and instead serves as an additional tool for the prosecution.” Robert G. Morvillo et al., Motion Denied: Systematic Impediments to White Collar Criminal Defendants’ Trial Preparation, 42 Am. Crim. L. Rev. 157, 160 n. 12 (2005).

• In expanding the discovery obligations under Federal Rule of Criminal Procedure 16 in 1974, the Criminal Rules Committee espoused the view that broad discovery contributes to the fair and efficient administration of criminal justice by providing the defendant with enough information to make an informed decision as to plea; by minimizing the undesirable effect of surprise at the trial; and by otherwise contributing to an accurate determination of the issue of guilt or innocence.

Fed. R. Crim. P. 16, Advisory Committee’s Note to the 1974 Amendment. The Federal Courts Committee agrees with that perspective and believes that the Criminal Rules Committee should apply the same philosophy in revising Rule 17. Doing so, the Federal Courts Committee believes, means using the materiality standard.

In addition, unlike the prosecution, which has a broad ability to collect documents from third parties through, among other things, grand jury subpoenas and search warrants, a Rule17(c) subpoena is a criminal defendant’s only vehicle to compel production of materials from third parties, including materials that are helpful or necessary to cross-examine the prosecution’s witnesses. A broader Rule 17(c) standard therefore strengthens the constitutional guarantee that provides criminal defendants “‘a meaningful opportunity to present a complete defense,’” Crane v. Kentucky, 476 U.S. 683, 690 (1986) (quoting California v. Trombetta 467
U.S. 479, 485 (1984)), which includes “the right to confront their accusers” through cross-examination. United States v. Tucker, 249 F.R.D. 58, 67 (S.D.N.Y. 2008). As then-District Judge Shira A. Scheindlin opined, the confrontation “right is meaningless if a defendant is denied the reasonable opportunity to obtain material evidence that could be crucial to that cross-examination.” Id.

While the Criminal Rules Committee’s rejection of the stricter Nixon admissibility standard is an important step in the right direction, the Federal Courts Committee does not believe that it goes far enough to safeguard a criminal defendant’s constitutional rights. For example, in Thor v. United States, 574 F.2d 215 (5th Cir. 1978), the defendant was unable to locate witnesses who he believed had exculpatory evidence and sought to subpoena an address book that he believed contained their addresses. Id. at 219. The Fifth Circuit upheld the district court’s denial of the subpoena on the ground that the address book was “not evidentiary” and therefore not available pursuant to a Rule 17(c) subpoena. Id. at 220. While the address book would meet a materiality standard, it would probably not meet the “likely to be admissible” standard proposed by the Criminal Rules Committee.

Similarly, the hypothetical set forth in the New York City Bar’s White Collar Committee’s letter provides another example in which the materiality and “likely to be admissible” standards may lead to divergent results. In the hypothetical, initially posed in Alan Silber and Lin Solomon, A Creative Approach for Obtaining Documentary Evidence from Third Parties (July 17, 2017), a defendant in an embezzlement prosecution believes that the financial data of a family business that encrypts its data using highly sophisticated software contains exculpatory evidence that would exonerate him. This hypothetical defendant filed a motion under Rule 17(c) to obtain both the data itself and the software program that would allow him to access the data. Even if the data itself meets the “likely to be admissible” standard, the necessary software program likely would not, with the result being that this defendant would not be able to access this evidence. Conversely, the hypothetical software program would satisfy a materiality standard.

Moreover, the Criminal Rules Committee’s stated concern in Section II.A.2 of the Report that a broader materiality standard may lead to “fishing expeditions” can be better addressed by retaining the limitation in current Rule 17(c) that affords courts the discretion to quash or modify subpoenas that are “unreasonable or oppressive.” As opposed to categorically preventing criminal defendants from accessing certain materials that the Federal Courts Committee believes are necessary to present a complete defense, adopting a broader Rule 17(c) standard while retaining the current language authorizing courts to quash or modify overbroad subpoenas promote the fair efficient administration of criminal justice, especially given the inherently fact-dependent nature of discovery issues.

Accordingly, the Federal Courts Committee believes that the Criminal Rules Committee should have instead adopted the materiality standard proposed by the New York City Bar’s While Collar Committee, or, alternatively, a “likely to lead to admissible evidence” standard, in clarifying and revising Nixon’s admissibility requirement.

We thank the Criminal Rules Committee for the opportunity to comment on the proposed changes.

II. Proposed Federal Rule of Evidence 707

The Federal Courts Committee also wants to highlight its support for the proposed Federal Rule of Evidence 707 (“Rule 707”) and suggests a few additional changes as set forth below. This proposed new rule would impose a standard for admission of “machine-generated” evidence, applying to that evidence the reliability and fit thresholds applied to expert testimony under Federal Rule of Evidence 702. Under the current rules, certain types of evidence generated by artificial intelligence programs can be admitted without an expert or other witness to vouch for (and be cross-examined on) the reliability of that evidence. While Federal Rule of Evidence 403 offers litigants the ability to object to such evidence, the proposed new Rule 707 provides greater clarity by utilizing the familiar Rule 702 standard and imposing a gatekeeper role on the Court.

The Federal Courts Committee respectfully wishes to raise two comments for consideration before the new Rule 707 is finalized. First, the Federal Courts Committee suggests including in Rule 707 a definition of “machine-generated evidence” to mean “evidence generated by a machine-based process or system to make predictions or draw inferences from existing data.” This definition, which adopts the language in the proposed Committee Note to the new Rule 707, more closely tracks the new rule’s purpose to ensure testing of potentially unreliable predictive and inferential software and systems, such as algorithms or machine-learning output. The Federal Courts Committee’s proposed definition also addresses concerns that the phrase “machine-generated evidence” is too broad because it would make clear that machine-generated materials that are not created by artificial intelligence are excluded. For example, the Federal Courts Committee’s proposed definition would exclude spreadsheets, emails, digital files of audio recordings like voicemails, computer-generated music in an audio
or video file, and documents created using a word processor (like this letter)—which are machine-generated but not typically used to make predictions or draw inferences from existing data.

Second, the Federal Courts Committee echoes the suggestion of the American Civil Liberties Union that the rule should be revised as follows (proposed additions underlined and proposed deletions crossed out): “When machine-generated evidence is offered without an expert witness and would be subject to Rule 702 if testified to by a witness, the court may admit the evidence only if it is limited to evidence not based on scientific, technical or other specialized knowledge or if expert testimony in support of the evidence satisfies the requirements of Rule 702(a)-(d). This rule does not apply to the output of simple scientific instruments.” The Federal Courts Committee suggests this revision for two reasons. First, it resolves concerns with unreliable methodologies and unreliable output by requiring an expert witness to testify in support of complex evidence whose reliability could not be established by a layperson operator or supervisor. Second, this revision clarifies the scope of applicable evidence by eliminating the vague phrase “simple scientific instruments” and instead harmonizing the proposed rule’s language with that of Rules 701 and 702 regarding “scientific, technical, or other specialized knowledge.” Enabling proponents of evidence that is not based on such knowledge to establish its admissibility without extensive proceedings, such as those that apply under Rule 702, should reduce the risk of overwhelming courts with the need to spend time on each submission of evidence, especially as machine learning is increasingly used to generate and supplement even the simplest kinds of information. For example, the Federal Courts Committee notes that summaries of voluminous evidence, permitted under Federal Rule of Evidence 1006, can often be better organized and prepared for presentation to a jury by a machine, without necessitating Federal Rule of Evidence 702-type inquiries into reliability.

The Federal Courts Committee also suggests that the Advisory Committee consider the application of the rule to evidence not offered for its truth.

The Federal Courts Committee applauds the Advisory Committee on Evidence Rules for its effort to establish rules governing this difficult technological area, especially given the speed with which machine learning and artificial intelligence technology have changed over time.

III. Proposed Changes to Federal Rules of Civil Procedure 26 and 45

The Federal Courts Committee also wants to highlight its support for the Advisory Committee on Civil Rules’ (the “Civil Rules Committee”) proposed amendments to Federal Rules of Civil Procedure 26 (“Rule 26”) and 45 (“Rule 45”).

The Civil Rules Committee has proposed a narrow, targeted amendment to Rule 26(a)(3)(A)(i) that would require parties’ pretrial disclosures to state whether any witness is expected to testify remotely, as opposed to in person. In practical terms, the amendment would require parties to identify in advance not only the witnesses they expect to present, but also the anticipated mode of that testimony. As the Civil Rules Committee stated in its report, this
proposal is linked to the companion amendments to Rule 45(c) (place of compliance) and Rule 45(b)(1) (service of subpoenas) and was prompted by the Ninth Circuit’s decision in In re
Kirkland, 75 F.4th 1030 (9th Cir. 2023).

The practical problems the proposed amendments seek to address arise from the combined operation of Rule 45(b)(1)’s personal service requirement and Rule 45(c)’s geographic limits when those provisions are applied to subpoenas seeking only remote testimony. Both provisions were developed to govern subpoenas compelling physical presence and travel, where territorial limits and personal service serve important due-process and fairness functions. When compliance is entirely remote, however, those concerns are not implicated.

The current interpretation of Rule 45(c) applied by some courts has produced results that are inconsistent with modern litigation practice. For example, under prevailing Ninth Circuit authority, a subpoena compelling remote testimony may be unenforceable absent personal service within the geographic limits of Rule 45(c), even where the witness would never be required to travel to the place of compliance. See In re Kirkland, 75 F.4th at 1042-47.

This result creates an unnecessary obstacle to remote depositions and testimony that does not meaningfully advance fairness or due process. The proposed amendment properly recognizes the distinction between remote and in-person compliance. It clarifies that territorial and service concepts designed for physical presence should not control subpoenas seeking remote testimony, which do not raise the same territorial or travel-based concerns and therefore should not be governed by service rules that no longer reflect modern practice.

The Federal Courts Committee believes that the amendment reflects contemporary litigation reality. Remote depositions and testimony are now routine, efficient, and widely accepted by courts and practitioners. The Federal Rules of Civil Procedure should reflect that reality. Remote testimony often reduces costs, minimizes witness burden, and increases access to evidence, particularly for non-party witnesses and parties of limited means. Where compliance occurs remotely, the traditional concerns underlying Rule 45(c)’s geographic limits are not implicated.

Importantly, existing protections under Rules 26(c) and 45(d) remain fully available to prevent undue burden, harassment, or abuse. Courts retain ample authority to limit, condition, or quash subpoenas where fairness concerns arise. The proposed amendment does not expand substantive subpoena power; it clarifies the permissible means of compliance.

From a practitioner’s perspective, the amendment to Rule 26 functions as a sensible coordination measure rather than a substantive expansion of discovery obligations. Requiring advance disclosure of anticipated remote testimony promotes transparency and fairness by giving parties and courts an opportunity to assess logistical and technological considerations, raise Federal Rule of Civil Procedure 43 or Rule 45 issues in a timely manner, and avoid last-minute motion practice.

For these reasons, the Federal Courts Committee supports the proposed amendments to Rules 26 and 45.

New York County Lawyers Association
Scott B. Klugman and Michael B. Eisenkraft
Co-Chairs, Committee on the Federal Courts
Aasiya Glover, Matthew Goldstein, Irina
Shpigel, and Hon. Joseph Kevin McKay (ret.)
Ad-Hoc Federal Rules Subcommittee

 

---

¹ This letter was approved for dissemination by NYCLA’s President as a Committee statement. This statement has not been approved by the NYCLA Board of Directors and does not necessarily represent the views of the Board.
² Available at https://www.uscourts.gov/sites/default/files/document/preliminary-draft-of-proposed-amendments-to-federal-rules_august2025.pdf.

 

***

This statement was approved for dissemination by NYCLA’s President as a Committee statement. This statement has not been approved by the NYCLA Board of Directors and does not necessarily represent the views of the Board.