“Whereas a human being might simply answer ‘I don’t know’ when faced with a question that exceeds their knowledge, AI models typically do not admit defeat.”  A. Spence and N. Bustard, Practical AI for Lawyers: Meeting the Challenges of Competence, Confidentiality, and Candor, 56 U. Balt. L.F. 331, 340 (2026).

CITATIONS TO HALLUCINATIONS BY ATTORNEYS

The leading Maryland decision on hallucinations is Mezu v. Mezu, 267 Md. App. 354, 367 (2025), cert. dismissed, 493 Md. 184 (2026).  It resulted in referral of the attorney to Bar Counsel.  There, the attorney tried to shift the blame to a law clerk. Maryland Fake Citation Case – Court Issues a Warning – Referral to Atty. Grievance Comm’n. (Nov. 2, 2025).  The Court wrote: “Before addressing the merits of the issues raised, we address this problem. We do so as a warning to others and to determine the court’s appropriate response in this case.” It explained:

This case… appears to be the first where the Maryland appellate courts have addressed the problems that occur when lawyers use AI, without the diligence required, in drafting briefs. The failure to use AI responsibly in legal research raises ethical issues and can result in sanctions when used improperly. It is unquestionably improper for an attorney to submit a brief with fake cases generated by AI.

The key facts applicable to sanctions were: “In determining whether to refer counsel to the Attorney Grievance Commission, we have considered the nature and severity of the conduct in this case. The brief here did not contain an isolated instance of a citation mistake. Rather, it cited multiple cases that did not exist, as well as others that did not support the proposition for which they were cited. When asked at oral argument about the concern when a lawyer obviously has not read the cases cited to the court, counsel indicated that this was not a unique circumstance, stating that he typically did not read the cases he cited in pleadings submitted to the court. Based on all the circumstances, we shall refer this case to the Attorney Grievance Commission.”

On January 14, 2026, the Court issued Boyd v. Lee, 2026 WL 111263 (Appl. Ct. Md. Jan. 14, 2026)(unreported), which is discussed in Hallucinations Are “Old News,” But a Recent Maryland Decision Addresses Ethical Contours  (Jan. 17, 2026).  There, an appeal was dismissed for procedural violations. The Court reiterated that:

It is unacceptable for counsel (or unrepresented parties, for that matter) to submit or attempt to rely on statements or authorities that are fabricated, hallucinated, or unsubstantiated, whether they are generated by artificial intelligence tools or “real” ones.  The signature of an attorney contained in a brief constitutes a certification that the attorney has read the brief, and that “to the best of the attorney’s knowledge, information, and belief there is good ground to support it[.]” Md. Rule 1-311(b). Thus, signing and filing a brief that contains hallucinated law may implicate the Maryland Attorneys’ Rules of Professional Conduct. Mezu v. Mezu, 267 Md. App. 354, 369, 374 (2025) (holding that counsel’s submission of brief with hallucinated case law implicated the Rules of Professional Conduct and referring case to the Attorney Grievance Commission). This conduct can also result in sanctions. Id. at 371-72 (declining to impose monetary sanctions when none were requested by opposing counsel).

“Though the court did not impose monetary sanctions against the three attorneys, it did issue public reprimands, disqualified them from the case, and referred them to the Alabama State Bar. Butler Snow narrowly escaped sanctions itself, but it was unable to escape the nationwide embarrassment, media ridicule, and client loss.”  Spence and Bustard, “Practical AI,” 56 U. Balt. L.F. at 338–39.

That was followed by Maryland Copy and Paste “Hallucinations” Case, citing Ajakaiye v. Oladipupo, 2026 WL 1733336 (Appl. Ct. Md. Jun. 16, 2026). There, “[t]he Court’s review of Father’s briefing confirms that, for most of the offending citations, Father’s counsel simply copied and pasted from the Father’s pro se second motion for reconsideration. Father’s counsel further concedes that he did not confirm whether AI was used in the Father’s preparation of that second motion for reconsideration, seemingly relying on the fact that the pro se briefing was drafted with the assistance of Father’s sister, who is a non-practicing attorney (having attended a foreign law school).”  Citing Mezu, the matter was referred to the Attorney Grievance Commission.

Next, in Benjamin v. State, 2026 WL 2491126 (App. Ct. Md. Aug. 25, 2026)(unreported), an appellant’s request to file a corrected brief three days before appellee’s briefing date was denied. “In this case, however, the brief was filed before the decision in Mezu, the inaccuracies were not excessive and did not include totally made up cases, counsel tried to correct the errors before the State filed its brief, and the State does not contend that it was harmed. Given these circumstances, we shall not, as we did in Mezu, refer counsel to the Attorney Grievance Commission. We again urge counsel, however, to use AI responsibly and check the accuracy of all citations before submitting a brief to this Court, or any other court.” [emphasis added]. For a discussion of this decision, please see Maryland: No Sanctions for Apparent Hallucinations (Aug. 29, 2026).

CITATIONS TO HALLUCINATIONS BY UNREPRESENTED LITIGANTS

The Dorsey decision involved an informal brief by an unrepresented litigant. Dorsey v. Livramento, 2026 WL 1413152 (Appl. Ct. Md. May 20, 2026)(unreported), cert. denied, 496 Md. 453 (2026).  After citing Mezu, the Court wrote:

Although there is no parallel disciplinary action for pro se litigants, Maryland Rule 1-311(c) authorizes us to strike the offending filing. Where, as here, the filing is an appellant’s principal brief, striking it could result in dismissal under Maryland Rule 8-602(c)(5) for failure to file a brief….

We exercise our discretion to decline striking Mother’s informal brief in its entirety. As explained below, several issues are not properly before us, so we need not identify or address the problematic citations because we do not reach the merits of these issues. See infra, Parts I–III. However, this does not excuse the impropriety of such conduct. To the extent an issue is properly before us, we will strike from the brief any cases that are fictitious and any authorities that clearly do not support the propositions for which they are cited. See infra, Part IV, notes 8–17. Mother is warned and put on notice that continued citation of fabricated cases or misrepresented legal authorities may result in sanctions, including the striking of future filings in their entirety.

In footnotes 8 through 17 parts of the brief were stricken.

Meziane v. Aitouche, 2026 WL 1101776 (Appl. Ct. Md. Apr. 23, 2026), involved a self-represented litigant whose brief contained two apparently fictitious citations. The Court cited Mezu and wrote:

[I]ndeed, attorneys who submit filings containing fake citations may be referred to the Attorney Grievance Commission depending on the nature and severity of the misconduct…. Although there is no parallel disciplinary action for pro se litigants, at the least, Rule 1-311(c) authorizes us to strike the offending filing. Where, as here, the filing is an appellant’s principal brief, striking it could result in dismissal under Rule 8-602(c)(5) for failure to file a brief….

That said, given the significant rights involved for the parties here, we shall exercise our discretion and take no action regarding Mother’s brief and will address the merits of her appeal. This does not excuse Mother’s suspected misconduct. Rather, despite her reliance on inapposite and apparently fictitious authorities, we can sufficiently address the issues raised.

This month, the Court wrote that “despite being made aware of its fictitiousness by the State’s opposition to that motion, [appellant] still included it in his brief on appeal.”  Sare v. Central Collection Unit, 2026 WL 3093192 (Appl. Ct. Md. Oct. 1, 2026)(unreported).  After citing Mezu, the Court wrote: “Although there is no parallel disciplinary action for pro se litigants, at the least, Rule 1-311(c) authorizes us to strike the offending filing. Where, as here, the filing is an appellant’s principal brief, striking it could result in dismissal under Rule 8-602(c)(5) for failure to file a brief….  Although we have decided not to take further action regarding Sare’s brief and will address the merits of his appeal, this does not excuse his suspected misconduct. We caution that future filings submitted to the Court with fake citations may be struck.”

In another recent decision concerning an unrepresented litigant, the Appellate Court of Maryland wrote:

In its brief, the Commission noted that Appellant cited to cases that did not exist and that he further cited an existing case for a proposition it did not support. Appellant did not address the Commission’s concern and in his reply brief once again cited the non-existent case law or propositions. In conducting review of this matter, we have confirmed that the authorities and proposition identified by the Commission are non-existent. While we recognize that Appellant is self-represented in this appeal, a note of caution is appropriate. We recently discussed the ethical issues concerning the use of AI to conduct legal research or generate filings, including the risk of “hallucinated” citations. See Mezu v. Mezu, 267 Md. App. 354, 367–68 (2025). Further, the citation of fake cases in a brief implicates multiple Maryland Rules. See, e.g., Md. Rule 1-311(b). Indeed, attorneys who submit filings with fake citations may face sanctions and may be referred to the Attorney Grievance Commission. See Mezu, 267 Md. App. at 367, 374. Although there is no parallel disciplinary action for pro se litigants, Md. Rule 1-311(c) permits us to strike an offending filing, which, in turn, could cause a dismissal under Md. Rule 8-602(c)(5) for failure to file a brief. In this case, we shall exercise our discretion and take no action regarding Appellant’s citation to non-existent cases or proposition and thus address the merits while noting that this does not excuse Appellant’s presentation of invalid authorities to this Court.

In the Matter of Gary Brown, Jr., 2026 WL 3025765 (Appl. Ct. Md. Oct. 2, 2026).

In each unreported decision, the Court wrote: “This is an unreported opinion. This opinion may not be cited as precedent within the rule of stare decisis. It may be cited for its persuasive value only if the citation conforms to Rule 1-104(a)(2)(B).”

Among other things, it is important for Maryland attorneys to be aware of:

• Proposed Changes to the MD Attorneys’ Rules of Professional Conduct Re: AI (Jul. 21, 2026);
• Maryland’s Proposed Anti-Hallucination Rule (Mar. 16, 2026);
• Child’s Best Interest Trumps Sanction for Possibly Hallucinated Case Citation (Mar. 15, 2026);
• “Hallucinations” by West’s CoCounsel? (Apr. 7, 2026);
• , Sanctions and Remedies Where Parties Submit Citations Hallucinated by Artificial Intelligence Programs—State Cases, 110 A.L.R.7th Art. 3 (2026).

Rule 104(b) is an old and familiar friend.  The federal Advisory Committee Notes track it back to 1972, stating:

In some situations, the relevancy of an item of evidence, in the large sense, depends upon the existence of a particular preliminary fact. Thus when a spoken statement is relied upon to prove notice to X, it is without probative value unless X heard it. Or if a letter purporting to be from Y is relied upon to establish an admission by him, it has no probative value unless Y wrote or authorized it. Relevance in this sense has been labelled “conditional relevancy.” Morgan, Basic Problems of Evidence 45-46 (1962). Problems arising in connection with it are to be distinguished from problems of logical relevancy, e.g., evidence in a murder case that accused on the day before purchased a weapon of the kind used in the killing, treated in Rule 401. [emphasis added].

But, it appears that “conditional relevance” will soon be no more.

The Committee on Rules of Practice and Procedure of the Judicial Conference of the United States published the “Preliminary Draft: Proposed Amendments to Federal Rules of Appellate, Bankruptcy, Civil, and Criminal Procedure, and the Federal Rules of Evidence” (August 2026).

The Report of the Advisory Committee on Evidence Rules describes the concept of conditional relevance as “unhelpful”  and “confusing.”  The proposed new language is double underlined below.  Removed language is stricken:

(a) In General. The court must decide any preliminary question about whether a witness is qualified, a privilege exists, or evidence is admissible. In so deciding, the court is not bound by evidence rules, except those on privilege. Unless these rules provide  otherwise, the proponent must establish the existence of a preliminary fact by a preponderance of the evidence.

10 (b) Relevance That Depends on a Fact. When the relevance of evidence depends on whether a fact exists, proof must be introduced sufficient to support a finding that the fact does exist. To establish that evidence is relevant, the proponent must demonstrate to the court that the trier of fact could reasonably find that the requirements of Rule 401 have been met. In making its relevance determination, the court may consider the evidence itself, other evidence, and common knowledge and experience. The court may admit the proposed evidence on the condition that the further proof be introduced later.

The end of conditional relevance is clear. The Committee Note states:

[T]he [proposed] amendment to Rule 104(b) eliminates the concept of “conditional relevance.” There is no reason for special treatment of facts that are conditionally relevant, because in almost all cases, the relevance of a particular fact is dependent on a showing of other facts and so the relevance is “conditional.” Moreover, if the existence of one fact is conditionally relevant upon another, the result is that both facts are “conditionally relevant.” See Ronald Allen, The Myth of Conditional Relevance, 25 Loyola L. Rev. 871-884, 49 879 (1992) (“Evidence is relevant only because there is an intermediate premise or set of premises that connects the evidence to some proposition involved in the litigation. But if determining the relevance of evidence always requires relying on some intermediate premise, no distinction can be drawn between relevancy and conditional relevancy.”).

The Committee Notes state:

Providing a uniform approach to all questions of relevance avoids confusion, especially because most courts currently use the same permissive approach to all relevant evidence, conditional or not. All questions of relevance should be governed by the same “a reasonable person could find” standard because the jury is in as good a position to determine relevance as is the court.

Subject to public comment and further review, the Committee also proposes other changes to Rule 104.  The proposal codifies the common-law requirement that preliminary questions of fact be determined by a preponderance of the evidence standard.

Further:

The [R]ule 104(b) test that “proof must be introduced sufficient to support a finding” has been revised, because a court’s preliminary determination of relevance is not limited to proof submitted to it. A court could find evidence relevant under Rule 104(b) by considering the evidence itself, as well as common knowledge and experience. For example, if the proffered evidence is that the defendant was carrying a gun, its relevance to a charge of armed robbery can be determined without any further proof. The standard has been changed to whether the factfinder “could reasonably find” the evidence to be relevant. That standard is the same as “sufficient to  support a finding.”

The Advisory Committee suggests that different geographic areas may have different “common knowledge and experience,” writing:

In deciding whether evidence is relevant, the judge must not only take into account the evidence formally in the record, but also the likely background knowledge and experience of a reasonable juror from the area where trial is being held. For example, suppose a trial concerns the cause of a large explosion and fire. The plaintiff claims that the defendant mishandled nitrogen fertilizer containing ammonium nitrate. A witness is asked whether such a fertilizer was stored on the premises and an objection on relevance is made. If the trial is in a rural state with a large farming population, the trial judge would likely overrule the objection because a reasonable person from that population would likely know the dangerous properties of the material.  If, by contrast, the trial were held in an urban area, the judge would likely conclude that evidence of the dangerous properties must be entered into the record to satisfy the requirement of relevance. [emphasid added].

The Committee states that the amended rule “retains the provision of allowing proof to be introduced later.”

Many other Rule changes are proposed in the lengthy report. One that will have wide-ranging impact is the proposal that blanket bans on use by unrepresented litigants of electronic filing systems are prohibited.

Maryland Rule 5-104(b) is similar to the current Fed.R.Evid. 104: “Relevance Conditioned on Fact. When the relevance of evidence depends upon the fulfillment of a condition of fact, the court shall admit it upon, or subject to, the introduction of evidence sufficient to support a finding by the trier of fact that the condition has been fulfilled.”

This blog was initially posted on  Electronic Discovery Reference Model and  JD Supra.

Mr. McIntyre sued the School District defendants seeking to enjoin them from stopping him from displaying his “Hate Has No Home Here” flag.  He anticipated filing for a preliminary injunction.

When it came to discovery, the McIntyre parties were at polar extremes.  Mr. McIntyre moved to expedite discovery in anticipation of a hearing on a motion for preliminary injunction.  The School District moved to stay discovery pending resolution of its Rule 12(b)(6) motion to dismiss.

THE RULES

Rule 26(d)(1) states: “A party may not seek discovery from any source before the parties have conferred as required by Rule 26(f), except in a proceeding exempted from initial disclosure under Rule 26(a)(1)(B), or when authorized by these rules, by stipulation, or by court order.”

Under Rule 26(f): “Except in a proceeding exempted from initial disclosure under Rule 26(a)(1)(B) or when the court orders otherwise, the parties must confer as soon as practicable–and in any event at least 21 days before a scheduling conference is to be held or a scheduling order is due under Rule 16(b).”  Rule 37(d) permits “delivery” of an early Rule 34 request.  Fed.R.Civ.P. 37(d)(2).

THE PLAINTIFF’S MOTION TO EXPEDITE DISCOVERY

The McIntyre court enunciated the standard governing a request to expedite discovery:

Expedited discovery may be granted upon a showing of good cause…. The party seeking expedited discovery bears the burden of showing good cause, which may be found where the need for expedited discovery, in consideration of the administration of justice, outweighs the prejudice to the responding party…. Furthermore, courts also consider whether the evidence may be lost or destroyed with time and whether the scope of the proposed discovery is narrowly tailored.

Id. at * 1 (cleaned up).

Plaintiff asserted that “courts have found good cause for expedited discovery when limited and narrowly tailored to a plaintiff’s requested injunctive relief,” and also acknowledged issues such as student privacy were implicated in his discovery requests.  He asserted that redaction could address those issues.  There were also other issues, such as privilege waiver.

The McIntyre court wrote: “Defendants, meanwhile, argue that expedited discovery is unnecessary because the responsive record is already before the Court.”  They pointed to policies in the court record and asserted that “they have already produced everything Plaintiff needs.”  Id. at *3.

“But, Plaintiff argues that this is not so, and points to several requested materials related to, but not identical to, the exhibits Defendants produced.”  Id.

Plaintiff prevailed:

The Court finds Plaintiff’s arguments persuasive. First, Defendants have not produced every document that Plaintiff requests, so discovery is not redundant. And, Defendants have not shown why depositions or document requests would prejudice them. Nor have they shown why this case is different from the numerous cases in this district where courts have permitted expedited discovery for factual matters relevant to pending motions for preliminary injunction.…  Indeed, in these situations, courts examine the discovery request, weighing the reasonableness of the request, along with the scope of the requested discovery.… The Court finds here that Plaintiff seeks expedited discovery specifically related to facts relevant to his Motion for Preliminary Injunction; further, he narrowly defines the discovery he seeks so that the request is not overly broad…. Defendants have failed to show that the request is unreasonable and have not pointed to any undue prejudice or hardship for them that outweighs Plaintiff’s need for the discovery.

Id. (cleaned up).

THE DEFENDANT’S MOTION TO STAY DISCOVERY

The McIntyre court then turned to the motion to stay discovery pending resolution of the motion to dismiss the Complaint.  The School District pointed to the court’s discretion to stay discovery pending resolution of potentially dispositive preliminary issues. Id. at *4. Defendants asserted that their Rule 12(b)(6) motion met that threshold standard.

The court wrote:

Meanwhile, courts in this district have followed two guiding principles when examining discovery stays: (1) as a general rule, courts have not been inclined to grant such motions; and (2) courts have imposed a “high bar” where a stay of discovery may be appropriate “only in special circumstances,” where the motion to dismiss is based on qualified or absolute immunity, or lack of jurisdiction.

Id. (citation omitted).

Mr. McIntyre responded that the School District’s motion was not based on “special circumstances,” such as immunity, jurisdiction, or limitations.

In reply, Defendants “argue that their 12(b)(6) motion does, in fact, present such threshold legal questions to warrant a stay of discovery.”  Id. at *4.  They pointed to a dispute over which constitutional framework governs and asserted that the resolution will impact discovery.

The court was unpersuaded.

Defendants’ Motion does not ask the Court to clarify the constitutional framework, but instead asks that the Court find one framework applies as a matter of law and dismiss the case. Thus, assuming that the Court denies the 12(b)(6) motion, that denial merely finds that Plaintiff has stated a claim for relief plausible on its face. Meanwhile, granting the motion would simply moot the discovery issue altogether. Accordingly, Defendants’ 12(b)(6) motion does not create the same sort of “threshold legal question” such as jurisdiction or immunity. Nor do Defendants show why this motion otherwise involves a special circumstance such that the motion to dismiss must be addressed before discovery proceeds.

The court also rejected a prematurity argument and addressed the requests under the State public records statute. It concluded:

Thus, the Court agrees with Plaintiff. As courts in this district have routinely found, a stay of discovery is not warranted where nothing about the pending motion appears to distinguish it from the garden-variety Rule 12(b)(6) motion which is typically insufficient to justify granting a stay of discovery…. Indeed, it is not enough that Defendants raise viable arguments for dismissal; rather, absent a threshold legal question, a stay is appropriate where it is patent that the case lacks merit…. While Defendants assert that their motion presents such threshold legal issues that warrant a stay…, they do not point to, nor can the Court find, any argument in the Motion to Dismiss that distinguishes it from a typical 12(b)(6) motion…. Nor does the Court find that the Complaint patently lacks merit…. And, lastly, Defendants have not shown that the burden of discovery prior to the resolution of the 12(b)(6) motion outweighs the burden to Plaintiff in litigating his Motion for Preliminary Injunction without discovery.

Id. at *5 (cleaned up); see also Motion to Dismiss Does Not Stay Discovery (Nov. 8, 2025); Should Discovery be Stayed Pending Ruling on Dispositive Motion? (Apr. 3, 2025); Discretionary Stay of Discovery Pending Decision on Dispositive Motion (Mar. 13, 2025).

CONCLUSION

The motion to expedite discovery was granted. The motion to stay discovery was denied.

In the related context of early requests to preserve ESI, see generally:

• A Request for a Preservation Order Was Denied: Preliminary Injunction Standard Was Not Applied (Apr. 3, 2026);
• Motion to Compel Preservation of RAM Denied Without Prejudice (Aug. 7, 2025);
• Failed Request for a Preservation Order (Jul. 13, 2022);
• Requests for a Preservation Order (Mar. 7, 2022).

This blog was initially posted on  Electronic Discovery Reference Model.

The court held that the information was work product and that the production and testimony had not waived work product protection under the precise terms of the ESI Protocol.

However, while not expressly stated, apparently after having reviewed the documents, the Aerosonic court also held that the recipient, Joby, had shown “substantial need and an inability to obtain the substantial equivalent….”  That showing overcame protection under the work product doctrine.

While not expressly stated, it seems to me that Aerosonic shows that a non-waiver and clawback agreement cannot “unring the bell.”  The court agreed that work product protection had not been waived.  However, the work product had been disclosed; the court concluded that the recipient had shown “substantial need” for it; and, therefore the otherwise-protected work product had to be produced. I wonder if the result would have been different if the work product had never been disclosed.

The documents at issue were:

(1) two deposition exhibits related to Aerosonic’s creation of alleged trade secrets for the purpose of suing Joby, (2) in depth testimony from multiple Aerosonic witnesses about those two deposition exhibits, and (3) Aerosonic witness testimony about a meeting between Aerosonic engineers to decide who should be listed as the purported creator of alleged secrets, and to compel production of additional related materials.

The first two categories, and some email, related to trade secrets and “critical dimensions.”

The ESI Protocol protected disclosed ESI, whether disclosure was inadvertent or otherwise.  It is not clear to me whether the protocol was incorporated into a Rule 502(d) order or, instead, was only a Rule 502(e) agreement between the parties.  Because this was in the Middle District of Florida, I am guessing that it was not an order.  Court Denied Unopposed Motions for Protective Order and Approval of ESI Protocol (Sep. 5, 2024).

The Aerosonic court wrote that an agreement may provide greater protections than those found in Rule 502.  2026 WL 900154, at *2.  Of course, an agreement among the parties binds only the parties, unlike a Rule 502(d) order.

First, the court determined that the disclosed “critical dimensions” information and related emails was work product because it was created at the direction of Aerosonic’s former attorney. The fact that the engineer meeting occurred was not protected.

Next, the court applied the ESI Protocol.

Aerosonic has not waived any privileges through inadvertently, or otherwise, producing the “critical dimensions” document and related email to Joby. Further, Aerosonic’s initial allowance of testimony as to the “critical dimensions” document before current counsel realized the document was protected by the work product doctrine was not a waiver of the protection. Upon discovering the circumstances of the document’s creation, Aerosonic requested to claw back the document and testimony concerning the document consistent with Rule 502 and the parties’ ESI protocol.

However, the “critical dimensions” document, the email, and the engineering meeting, were within the scope of discovery.  That led to the following.

The Aerosonic court then applied the “substantial need” prong of the work product doctrine:

Joby also has shown it has a substantial need and an inability to obtain the substantial equivalent of the “critical dimensions” document and the testimony concerning the same document (and related email) and the engineering meeting without undue hardship. Moreover, the discovery of the disputed documents and testimony related to those documents is not a fishing expedition….  Specifically, Joby has overcome the qualified work product protection by demonstrating a substantial need to obtain (1) the presentation containing an image of Joby’s probe presented by Mr. Van Bemden in the meeting, (2) all photos, videos, and other visual images of Joby’s air data probe presented by Mr. Van Bemden at the meeting, and (3) all notes taken by the attendees of the meeting during the meeting, except any notes that reveal an attorney’s mental impressions, opinions, or legal theories. Therefore, Aerosonic is compelled to produce these documents.

The court held that “Joby may retain possession of the ‘critical dimensions’ document and related email Aerosonic has already disclosed and may continue to inquire at the remaining depositions into the factual information concerning such documents and the engineering meeting. Aerosonic must produce the documents described in the three categories listed in the previous paragraph because Joby has made a sufficient showing to overcome the qualified fact work protection of these documents.”

However: “To the extent Joby seeks a finding of a broad waiver of privilege or to further compel any additional documents, the motion is DENIED. Further, this Order should not be interpreted to authorize Joby to inquire into Aerosonic’s prior counsel’s litigation strategy.”

Aerosonic was decided on an emergency basis, one day before the discovery deadline. Id. at *1. In this blog, I am assuming that the contents of the disclosed information was known to the court.  Fed.R.Civ.P. 26(b)(5)(B) states: “After being notified, a party must promptly return, sequester, or destroy the specified information and any copies it has; must not use or disclose the information until the claim is resolved; must take reasonable steps to retrieve the information if the party disclosed it before being notified; and may promptly present the information to the court under seal for a determination of the claim.” [emphasis added].

This blog was initially posted on  Electronic Discovery Reference Model.

Maryland’s Standing Committee on Rules of Practice and Procedure has recommended an amendment to Maryland Rule 1-311 to address citation of “hallucinated” or fabricated cases.  Maryland’s Proposed Anti-Hallucination Rule (Mar. 16, 2026).  That issue has also been addressed in two recent decisions.  Maryland Fake Citation Case – Court Issues a Warning – Referral to Atty. Grievance Comm’n. (Nov. 2, 2025); Hallucinations Are “Old News,” But a Recent Maryland Decision Addresses Ethical Contours (Jan. 17, 2026).

This blog focuses on a different issue. Maryland has long had a “computer-generated evidence” rule applicable to civil cases, Md. Rule 2-504.3.

“Just as the telegraph gave way to the telephone, the stagecoach gave way to the automobile, and the typewriter gave way to the word processor, so too will courtroom chalkboards, easels and blow-up placard charts give way to computer-generated exhibits.” Galves, 13 Harv. J. L. & Tech. at 300. In fact, some jurisdictions have amended their rules of civil procedure to facilitate the use of CGE’s in the courtroom. See e.g., Md. Rules 2-504.3.”

Com. v. Serge, 58 Pa. D. & C.4th 52, 84 (Com. Pl. 2001)(emphasis added).

However, since it was adopted in 1998 and amended in 2009, no reported Maryland case has cited Rule 2-504.3. The rule “sets out procedures regarding the disclosure, discovery, and pretrial challenges applicable to computer animations or simulations.”  The Honorable Paul W. Grimm (ret.), Claudia Diamond, “Low-Tech Solutions to High-Tech Wizardry: Computer-Generated Evidence,” Md. B.J., July/August 2004, at 5, 6.

Rule 2-504.3 is a procedural rule; it does not address the substantive admissibility of such evidence.”  Id. at *10. It “provides comprehensive guidance to the bench and bar regarding the use of animations or simulations as evidence at trial and the trial court’s handling of any dispute about such evidence.

Judge Grimm and Ms. Diamond summarized its operation.  First, notice is required.  Then an objection may be made.

The filing of an objection requires the court to hold a hearing to determine the admissibility of the evidence. The rule also permits the trial judge to appoint an expert to assist it in ruling on the objection and to assess costs and expenses to one or more of the parties.

If the evidence is ruled admissible, the party opposing it need not renew objections again at trial, as the earlier hearing preserves them for appeal. Likewise, if the evidence is excluded, the issue is preserved for appeal, and no objection at trial is needed.

The procedures of Rule 2-504.3 allow practitioners, as well as the bench, to take the time and expend the necessary effort to resolve what may be complex evidentiary issues deliberately– rather than hastily addressing them in the midst of trial.

Id. at *11.

In my opinion, the current Rule likely applies to evidence created using GenAI.  This blog recommends an amendment to Maryland Rule 2-504.3 to make that explicit and to address what Maura Grossman, J.D., Ph.D., and the Hon. Paul W. Grimm (ret.) call “acknowledged” Generative A.I. evidence.  See “Judicial Approaches to Acknowledged and Unacknowledged AI-Generated Evidence,” 26 Col. Sci. & Tech. L. Rev. 110 (2025).

This suggestion borrows from a Maryland criminal statute addressing facial recognition technology, a form of AI. Facial Recognition Technology in Maryland Criminal Cases (Jan. 28, 2026); Maryland’s Facial Recognition Technology Statute (Aug. 14, 2025); Criminal Conviction Reversed After State Failed to Timely & Fully Disclose its Use of a Type of Artificial Intelligence (Aug. 11, 2025).

Further, this suggestion borrows from  “Artificial Intelligence (AI) and the Practice of Law,” by the Hon. Xavier Rodriguez, which is available at 24 Sedona Conf. J. 783 (2023).  The Sedona Conference article suggests that rules of procedure should ensure that adequate deadlines are set for a Daubert hearing. Id. at 801.

Judge Rodriguez continues: “Further, a pretrial hearing will likely be required for a trial court to assess the degree of accuracy with which the AI system ‘correctly measures what it purports to measure’ or otherwise ‘demonstrates its validity and reliability.’” Id. at 803. Judge Rodriguez wrote that “AI evidence may require that the offering party disclose any training data used by the AI platform to generate the exhibit.”  Id. at 802, as quoted in A Review of Sedona’s “Artificial Intelligence (AI) and the Practice of Law” by The Hon. Xavier Rodriguez (Sep. 27, 2023).  This proposed amendment to the Maryland Rule calls for notice, disclosure, discovery, and provides for a pretrial hearing if there are objections.

This suggestion is not a universal solution of all issues associated with GenAI.  It will be most effective where the proponent “plays by the rules,” i.e., with “acknowledged” GenAI.  By its terms, it would also address, and likely preclude admission of, “unacknowledged” GenAI evidence because failure to give notice would definitionally violate its terms. However, candidly, it would have less effect on the undisclosed GenAI, such as deepfakes, due to the difficulty in determining that “unacknowledged” GenAI is in fact AI-generated.

However, I suggest that this suggested amendment would be beneficial in addressing “acknowledged” GenAI.  The fact that it does not solve every problem does not undermine its value.

Rule 2-504.3 addresses “Computer-Generated Evidence.” This suggestion clarifies it to expressly encompass evidence created by GenAI.  Its key features are that it requires prior notice of the use of a GenAI exhibit; disclosure; discovery; a pre-trial hearing; and, a court-appointed expert.

Proposed additional language to the current Rule is bold, italicized, and underlined.  Minor deletions are marked as strikeouts.

RULE 2-504.3. COMPUTER-GENERATED EVIDENCE

(a) Definition–Computer-Generated Evidence. “Computer-generated evidence” means: (1) a computer-generated aural, visual, or other sensory depiction of an event or thing; and (2) a conclusion in aural, visual, or other sensory form formulated by a computer program or model, including as to both (1) and (2), without limitation, an image, video or animation, or audio recording that was created using Generative Artificial Intelligence. The term does not encompass photographs merely because they were taken by a camera that contains a computer; documents merely because they were generated on a word or text processor; business, personal, or other records or documents admissible under Rule 5-803 (b) merely because they were generated by computer; or summary evidence admissible under Rule 5-1006, spread sheets, or other documents merely presenting or graphically depicting data taken directly from business, public, or other records admissible under Rules 5-802.1 through 5-804.

(b) Notice: Use at Trial –

(1) Except as provided in subsection (b)(2) of this Rule, any party who intends to use computer-generated evidence at trial for any purpose shall file a written notice within the time provided in the scheduling order or no later than 90 days before trial if there is no scheduling order that:

(A) contains a descriptive summary of the computer-generated evidence the party intends to use, including (i) a statement as to whether the computer-generated evidence intended to be used is in the category described in subsection (a)(1) or subsection (a)(2) of this Rule, (ii) a description of the subject matter of the computer-generated evidence, and (iii) a statement of what the computer-generated evidence purports to prove or illustrate; and

(B) is accompanied by a written undertaking that the party will take all steps necessary to (i) make available any equipment or other facility needed to present the evidence in court, (ii) preserve the computer-generated evidence and furnish it to the clerk in a manner suitable for transmittal as a part of the record on appeal, and (iii) comply with any request by an appellate court for presentation of the computer-generated evidence to that court.

(2) Any party who intends to use computer-generated evidence at trial for purposes of impeachment or rebuttal shall file, as soon as practicable, the notice required by subsection (b)(1) of this Rule, except that the notice is not required if computer-generated evidence prepared by or on behalf of a party-opponent will be used by a party only for impeachment of other evidence introduced by that party-opponent. In addition, the notice is not required if computer-generated evidence prepared by or on behalf of a party-opponent will be used only as a statement by a party-opponent admissible under Rule 5-803 (a).

(c) Notice: Use in Motions

(1) Any party who intends to use computer-generated evidence in a motion or response to a motion for any purpose shall file a written notice with the motion or response that:

(A) contains a descriptive summary of the computer-generated evidence the party intends to use, including (i) a statement as to whether the computer-generated evidence intended to be used is in the category described in subsection (a)(1) or subsection (a)(2) of this Rule, and, (ii) a statement of what the computer-generated evidence purports to prove or illustrate; and

(B) is accompanied by a written undertaking that the party will take all steps necessary to (i) make available any equipment or other facility needed to present the evidence in court, (ii) preserve the computer-generated evidence and furnish it to the clerk in a manner suitable for transmittal as a part of the record on appeal, and (iii) comply with any request by an appellate court for presentation of the computer-generated evidence to that court.

(C) if the computer-generated evidence is to be used in a motion, a responding party shall be permitted sufficient time under section (d) of this Rule for additional discovery relating to the admissibility of the computer-generated evidence.  If the computer-generated evidence is to be used in a response to a motion, the moving party shall be permitted sufficient time under section (d) of the Rule for additional discovery relating to the admissibility of the computer-generated evidence.

(c) (d) Required Disclosure; Additional Discovery. Within five days after service of a notice under section (b) or (c) of this Rule, the proponent shall make the original or an accurate duplicate of the computer-generated evidence available to any party. Notwithstanding any provision of the scheduling order to the contrary, the filing of a notice of intention to use computer-generated evidence entitles any other party to a reasonable period of time to discover any relevant information needed to oppose the use of the computer-generated evidence before the court holds the hearing provided for in section (e) (f)of this Rule.

(d) (e) Objection. Not later than 6021 days after service of a notice under section (b)  or (c) of this Rule, a party may file any then-available objection that the party has to the use at trial of the computer-generated evidence and shall file any objection that is based upon an assertion that the computer-generated evidence does not meet the requirements of Rule 5-901(b)(9).[1] An objection based on the alleged failure to meet the requirements of Rule 5-901(b)(9)[2] is waived if not so filed, unless the court for good cause orders otherwise.

(e) (f) Hearing and Order. If an objection is filed under section (d)(e) of this Rule, the court shall hold a pretrial hearing on the objection. If the hearing is an evidentiary hearing, the court may appoint an expert to assist the court in ruling on the objection and may assess against one or more parties the reasonable fees and expenses of the expert. In ruling on the objection, the court may require modification of the computer-generated evidence and may impose conditions relating to its use at trial. The court’s ruling on the objection shall control the subsequent course of the action. If the court rules that the computer-generated evidence may be used at trial, when it is used, (1) any party may, but need not, present any admissible evidence that was presented at the hearing on the objection, and (2) the party objecting to the evidence is not required to re-state an objection made in writing or at the hearing in order to preserve that objection for appeal. If the court excludes or restricts the use of computer-generated evidence, the proponent need not make a subsequent offer of proof in order to preserve that ruling for appeal.

(f) (g) Preservation of Computer-Generated Evidence. A party who offers or uses computer-generated evidence at any proceeding shall preserve the computer-generated evidence, furnish it to the clerk in a manner suitable for transmittal as a part of the record on appeal, and present the computer-generated evidence to an appellate court if the court so requests.

Committee note: This section requires the proponent of computer-generated evidence to reduce the computer-generated evidence to a medium that allows review on appeal. The medium used will depend upon the nature of the computer-generated evidence and the technology available for preservation of that computer-generated evidence. No special arrangements are needed for preservation of computer-generated evidence that is presented on paper or through spoken words. Ordinarily, the use of technology that is in common use by the general public at the time of the hearing or trial will suffice for preservation of other computer- generated evidence. However, when the computer-generated evidence involves the creation of a three-dimensional image or is perceived through a sense other than sight or hearing, the proponent of the computer-generated evidence must make other arrangements for preservation of the computer-generated evidence and any subsequent presentation of it that may be required by an appellate court.

To be clear, this is my suggestion and it is not a current proposal to amend the Rule.

This blog was initially posted on  Electronic Discovery Reference Model and  JD Supra.

____

[1] Rule 5-901(b)(9) addresses authentication by: “Evidence describing a process or system used to produce the proffered exhibit or testimony and showing that the process or system produces an accurate result.”  It may be preferable to cite the authentication rule, Rule 5-901, in its entirety.

[2] Id.

Md. Rule 1-311 is the analog of Fed.R.Civ.P.26(g). As currently drafted, Rule 1-311 states that the signature of an attorney on a pleading or paper constitutes a certification that the attorney has read it.  The proposed amendment will add that the attorney has confirmed the existence and authenticity of each cited legal authority:

The signature of an attorney on a pleading or paper constitutes a certification that the attorney has read the pleading or paper and has confirmed the existence and authenticity of each legal authority cited; that to the best of the attorney’s knowledge, information, and belief there is good ground to support it the pleading or paper; and that it the pleading or paper is not interposed for improper purpose or delay.

Willful violations can result in attorney disciplinary action.

The Reporter’s Note cites to Mezu v. Mezu, 267 Md. App. 354 (2025). That, and other, Maryland hallucination cases were described in:

• Maryland Fake Citation Case – Court Issues a Warning – Referral to Atty. Grievance Comm’n. (Nov. 2, 2025);
• Hallucinations Are “Old News,” But a Recent Maryland Decision Addresses Ethical Contours (Jan. 17, 2026)

See generally, Child’s Best Interest Trumps Sanction for Possibly Hallucinated Case Citation (Mar. 15, 2026); Fourth Circuit Technology Agnostic Hallucination Decision Results in Admonishment (Mar. 13, 2026) Litigant Forfeited Claim that Court Erred by Relying on Hallucinated/Fictional Cases (Mar. 11, 2026).

The Reporter’s Note states that the recommended amendment will “make explicit what already is implicit in the Rule: an attorney has the obligation to confirm the existence and authenticity of any cases cited in a filing.”

Hallucinated case citations are “old news.”  To paraphrase, “[I]t is no longer amateur hour.” (Feb. 12, 2021)(citation omitted).

Maryland has not been immune.  Maryland Fake Citation Case – Court Issues a Warning – Referral to Atty. Grievance Comm’n. (Nov. 2, 2025)(Court wrote that: “Implicit in the requirement that an attorney certify that there is good ground to support the contentions raised is that the attorney has read the primary cases on which the attorney relies for the argument submitted.”).

In Boyd, Mr. Boyd appealed from dismissal of his Complaint and denial of subsequent motions.  The Appellate Court of Maryland dismissed his appeal.

Mr. Boyd’s Complaint for divorce on the grounds of cruelty was denied after a trial in which he appeared pro se.  He was represented by counsel on appeal.  His appellate counsel filed a brief and purported “record extract.”  The latter was deemed inadequate.  There were also other procedural deficiencies—insufficient copies were filed—and the Court issued a show cause order to which there was no response.

The next event was a sua sponte order stating: “On December 2, 2025, we ordered Mr. Troy to show cause, in writing, why he should not be sanctioned and why he should not be referred to the Attorney Grievance Commission for citing to ‘hallucinated’ and unsubstantiated case law in Mr. Boyd’s brief.”  Id. at *2 (emphasis added).  The Court cited six irregularities and later added two more to that list.

The appeal was dismissed for failure to respond to the order to file sufficient paper copies and otherwise respond to the show cause orders. Id. at *2.  The Court added:

We briefly address the evident artificial intelligence issue that, while not dispositive here, warrants comment. It is unacceptable for counsel (or unrepresented parties, for that matter) to submit or attempt to rely on statements or authorities that are fabricated, hallucinated, or unsubstantiated, whether they are generated by artificial intelligence tools or “real” ones.  The signature of an attorney contained in a brief constitutes a certification that the attorney has read the brief, and that “to the best of the attorney’s knowledge, information, and belief there is good ground to support it[.]” Md. Rule 1-311(b). Thus, signing and filing a brief that contains hallucinated law may implicate the Maryland Attorneys’ Rules of Professional Conduct. Mezu v. Mezu, 267 Md. App. 354, 369, 374 (2025) (holding that counsel’s submission of brief with hallucinated case law implicated the Rules of Professional Conduct and referring case to the Attorney Grievance Commission). This conduct can also result in sanctions. Id. at 371-72 (declining to impose monetary sanctions when none were requested by opposing counsel).

Id. at *2 (emphasis added).

It is noteworthy that the Court relied on the certification requirement. The blog Does Maryland Have an Analog to Fed.R.Civ.P. 26(g)? (Dec. 23, 2022) discusses the Hon. Paul W. Grimm’s (ret.) article suggesting that Md. Rule 1-311, “when read in tandem with Rule 1-341, functions similarly to Fed. R. Civ. P. 26(g),” in some respects, but requires a higher standard in others.

In Boyd, the Appellate Court also, in the alternative, reached the merits. Id. at *3.  It then concluded: “Despite being given multiple opportunities, Mr. Troy failed to file with this Court eight paper copies of Mr. Boyd’s brief and a record extract. See Md. Rules 20-404(b) and 20-406(a)(2). We, accordingly, exercise our discretion pursuant to Maryland Rule 8-602(c)(5) and dismiss the instant appeal.”

The Court wrote: “This is an unreported opinion. This opinion may not be cited as precedent within the rule of stare decisis. It may be cited for its persuasive value only if the citation conforms to Maryland Rule 1-104(a)(2)(B).”

For reviews of recent books on Artificial Intelligence, see Book Review: Craig Ball, “The Leery Lawyer’s Guide to AI and LLMs in Trial Practice” (Jan. 16, 2026); Book Review: John Tredennick and William Webber, “Generative AI for Smart Discovery Professionals” (Jan. 15, 2026); Book Review: Jim Sullivan, “The Book on AI Doc Review” (Jan. 13, 2026); Book Review: Tom O’Connor, “Artificial Intelligence for the Rest of Us” (Jan. 12, 2026).

For additional information on A.I., click on the “TAG” named “Artificial Intelligence” in the drop down menu at the top of this blog. For example a recent blog discusses Artificial Intelligence in Maryland Courts (Aug. 24, 2025).

UPDATE: See Pamela Langham, U.S. District Court of Maryland Addresses Hallucinated Cases | Maryland State Bar Association (Jan. 7, 2026). In Lafferty v. Theiss, No. 1:2024cv02642 – Document 29 (D. Md. 2025), Ms. Langham wrote that: “Plaintiff’s counsel further represented that both counsel and the firm had taken concrete steps to prevent overreliance on AI in future pleadings. To avoid recurrence of the errors, the firm instituted several protocols. First, every quotation or parenthetical citation must now be substantiated by a PDF copy of the underlying judicial opinion, which is to be downloaded and thoroughly reviewed by the attorney responsible for the pleading. Additional measures include requiring each checker to independently verify all citations and quotations prior to filing, confirming the accuracy of citations through Shepardizing or KeyCite, retaining the verified authorities for potential court review, and mandating a secondary audit of all filings by another member of the firm to ensure citation accuracy.”

Ms. Langham reported that, in Neal et. al v. Frayer et. al., No. 8:2024cv00778 – Document 31 (D. Md. 2025): “The court noted that plaintiff’s counsel had faced similar errors in a separate case, Lafferty v. Theiss. Despite the mistakes and errors in plaintiff’s counsel’s response in the Neal case, the court decided not to issue a show cause order or impose sanctions, citing counsel’s prior assurances in the Lafferty case that corrective measures had been implemented at his firm to prevent future errors.”

First, prior to 2000, Rule 26(b)(1) read: “The information sought need not be admissible at the trial if the information sought appears reasonably calculated to lead to the discovery of admissible evidence.” The intent was to make it clear that relevant information could not be withheld from discovery because it was, for example, hearsay or otherwise inadmissible.  Garcia v. State Farm Mutual Automobile Ins. Co., 2025 WL 2966364, at n. 7 (D.N.M. Oct. 21, 2025)(citation omitted).

Second, in 2000, the Rules were amended to state: “Relevant information need not be admissible at the trial if discovery appears reasonably calculated to lead to the discovery of admissible evidence.” This was a “housekeeping amendment.”  It clarified that inadmissible evidence must be relevant to be discoverable. Garcia, 2025 WL 2966364, at *4.

Third, in 2015, the phrase “if discovery appears reasonably calculated to lead to the discovery of admissible evidence” was deleted.[1]

Frequently, that change is overlooked.  See Fischer v. Forrest, 2017 WL 773694, at *1 (S.D.N.Y. Feb. 28, 2017)(Peck, J.) )(“Despite this clear change, many courts [and lawyers] continue to use the phrase. Old habits die hard…. The test going forward is whether evidence is ‘relevant to any party’s claim or defense,’ not whether it is ‘reasonably calculated to lead to admissible evidence.’”); A Wake Up Call Revisited: Read the Rules & Don’t Argue “Not Reasonably Calculated” in Federal Courts (Feb. 25, 2025; “Reasonably Calculated to Lead to Discovery of Admissible Evidence” (Nov. 6, 2024).

As of 2015, Rule 26(b)(1) reads: “Information within this scope of discovery need not be admissible in evidence to be discoverable.”  In my words, the emphasis focuses on the “scope of discovery,” i.e., the claims, defenses, and proportionality.

Garcia describes those amendments and explains them.

The Garcia court wrote that deletion of the “reasonably calculated” phrase “did not necessarily change discovery’s scope, but clarified it.” 2025 WL 2966364, at *4. The deleted phrase had “been used by some, incorrectly, to define the scope of discovery.” Id. (quoting the Advisory Committee). The Garcia court explained that:

Despite the 2000 amendment’s attempt to correct that misapplication by adding the word “Relevant” at the beginning of the sentence, the Committee found that the “reasonably calculated” phrase had “continued to create problems.”

Id.

In Garcia, the plaintiff referred to the “reasonably calculated” language.  Defendant State Farm argued that the language was superseded and: “In making his legal argument, Plaintiff does not adhere to these newer discovery standards; rather, he focuses on the old, broader discovery standards.” Id. at *4.

The Garcia court disagreed. Id.  It explained:

Notably, discovery of nonprivileged information not admissible in evidence remains available so long as it is otherwise within the scope of discovery.... In other words, despite the 2015 changes—made for the purpose of clarifying and correcting past misuse of the deleted language—relevance is still to be construed broadly to encompass any matter that bears on, or that reasonably could lead to other matter that could bear on any party’s claim or defense.... In short, Plaintiff was not wrong in referencing the “reasonably could lead to” language, although confusion generated by that language probably warrants its avoidance.

Id. at *5 (cleaned up; quotations and citations omitted).

The Garcia court stated “that the 2015 deletion did not change the basic principle that information does not have to be admissible in order to be subject to discovery.” Id. (emphasis in original).

In its exhaustive opinion, the Garcia court resolved more discovery disputes than I can count. In one holding, it wrote: “In sum, the discovery requests, as limited by Plaintiff in his reply brief appear relevant on their face—and may lead to the discovery of admissible evidence regarding the scope and the nature of investigation, if any, that State Farm undertook prior to rejecting Plaintiff’s UIM/UM claim.”  2025 WL 2966364, at *12 (emphasis added).

In note 8, the Garcia court cited Judge Browning’s observation in Landry v. Swire Oilfield Servs., L.L.C., 323 F.R.D. 360, 379 (D.N.M. 2018), that the deletion did not change the scope of discovery and “[a]rguably, older lawyers will have to learn a new vocabulary and ignore the one they have used for decades. If the changes were not made to change the scope of discovery, it is unclear what the benefit of all this really is.”

This blog was initially posted on  Electronic Discovery Reference Model.

____

[1] There were other important changes in 2015 that are not discussed here.

THE RULES GOVERNING UNPUBLISHED DECISIONS

The People of the State of California opinion states: “This disposition is not appropriate for publication and is not precedent except as provided by Ninth Circuit Rule 36-3.”

However, Fed.R.App.P. 32.1 states that “[a] court may not prohibit or restrict the citation of federal judicial opinions, orders, judgments, or other written dispositions that have been: (i) designated as ‘unpublished,’ ‘not for publication,’ ‘non-precedential,’ ‘not precedent,’ or the like; and (ii) issued on or after January 1, 2007.”

The 2006 Advisory Committee Note states that the Rule 32.1 “says nothing about what effect a court must give to one of its unpublished opinions….”  The Committee Note adds: “Under Rule 32.1(a), a court of appeals may not prohibit a party from citing an unpublished opinion of a federal court for its persuasive value or for any other reason.” [emphasis added].

By way of comparison, Maryland State courts have taken a different approach to unreported decisions.  Elimination of Judicial Disappearing Ink in Maryland (May 6, 2023); Maryland’s New and Improved Unreported Opinion Rule Does Not Go Far Enough (Jun. 1, 2024).

FACTS AND PROCEDURAL POSTURE

While the People of the State of California mandamus petition named the District Court as a respondent,[1] the litigation at issue involves Meta Platforms, Inc., and others as the real parties in interest.  The underlying claims were summarized by the court:

A coalition of State Attorneys General (the State AGs) have brought an enforcement action against Meta Platforms, Inc., the parent company of Facebook, under various state consumer-protection statutes and the Children’s Online Privacy Protection Act of 1998, 15 U.S.C. § 6501 et seq. The State AGs allege that Meta designed intentionally addictive products that harm teenagers’ mental health.

Id. at *1.

“On February 23, 2024, Meta sent the State Attorneys General a list of state agencies ‘that may possess information relevant to the claims or defenses.’ …  Meta served requests for production of documents under Rule 34 on the State Attorneys General on February 27, 2024.”  In Re: Social Media Adolescent Addiction/Personal Injury Products Liability Litigation, 2024 WL 4125618, at *1 (N.D. Cal. Sep. 6, 2024).

Then, in discovery, Meta sought nonparty State agency documents using a request for production of documents pursuant to Fed.R.Civ.P. 34, “arguing that the documents are within the control of the State AGs.”  The State AGs objected, asserting that they lack “legal control over independent state agency documents.” 2025 WL 2427608, at *1. The U.S. Magistrate Judge rejected those assertions and ordered that the State AGs produce the requested documents. Id. The District Judge agreed in large part, but overruled the Magistrate Judge in part.

In the words of the Magistrate Judge:

Each State Plaintiff is represented by attorneys from the office of their respective State Attorney General. After Meta served requests for production of documents under Federal Rule of Civil Procedure 34 on the State Attorneys General, the Parties disputed whether or not certain identified state agencies should be subject to party discovery, and thus, whether the States Attorneys General should not only collect and produce documents from the Attorney General’s office but also from relevant custodians within the identified state agencies…. [The State AGs] object to treating their respective state agencies as being subject to party discovery and insist that all of these agencies are third parties from whom Meta should seek documents by subpoenas under Federal Rule of Civil Procedure 45.

In re Soc. Media Adolescent Addiction, 2024 WL 4125618, at *1 (emphasis added).

The State of California and other State AGs petitioned for a writ of mandamus ordering the district court to grant their objections.  2025 WL 2427608, at *1. The Ninth Circuit granted that Petition as to California and several other states.  The mandamus petition was denied as to one State and dismissed as moot to others.

THE NINTH CIRCUIT’S HOLDINGS

The Ninth Circuit provided a non-exhaustive list of five factors to consider on a mandamus petition, and added that the lack of “clear error” “will always defeat a petition for mandamus….” 2025 WL 2427608, at *1 (citation omitted). Conversely, clear error could support mandamus.

The Ninth Circuit agreed with the State AGs “that the discovery orders were clearly erroneous because they failed to require a showing of the existence of an attorney-client relationship between the Attorneys General and nonparty state agencies.”  Id. at *2.  The Ninth Circuit stated that the district court orders “erroneously inferred the future existence of such a relationship….”  Id.

Two important principles were stated:

• “Without an affirmative showing that an attorney-client relationship exists [between an AG and an agency], discovery of nonparty state agency documents should be conducted under Rule 45.” Id. at *2 (emphasis added).
• “Insofar as the parties litigated this case at the extremes, we reject both the view that a state attorney general never has control over the documents of a nonparty state agency and the view that a state attorney general always has control over the documents of a nonparty state agency.” Id. at *2 n. 3 (emphasis in original).

In short, “control” was a question of fact and here (with one exception) Meta did not provide a sufficient factual predicate to show control by the State AGs over the State agencies.

Another important holding was that, even though the Ninth Circuit was considering objections to discovery under the Federal Rules of Civil Procedure, “we must examine state law to determine whether the attorney general of a state controls the documents of nonparty state agencies.”  Id. at *2 (emphasis added).  Applying that principle, the Ninth Circuit found “clear error” in the district court failing to give People ex rel. Lockyer v. Superior Court, 19 Cal. Rptr. 3d 324 (Cal. Ct. App. 2004), “controlling weight.” Id.  Simply stated, under Lockyer, when the People prosecute an enforcement action, they do not have possession, custody, or control over documents of any state agency.  Id.[2]

In addition to “clear error,” the Ninth Circuit held that the State AGs satisfied the other requisites of mandamus.  It stated that high-level government officials need not incur sanctions, such as contempt, in order to seek relief.  And, it stated that mandamus is appropriate to address important, novel discovery and procedural issues. Id. at *2.  Further, while discovery costs do not generally support mandamus, here California “has represented that complying with the discovery orders has required the expenditure of millions of dollars in public funds and ‘strained’ the ability of California state agencies ‘to perform essential public functions.’”  Id.

The Ninth Circuit held that mandamus was proper for California and the Attorneys General of Arizona, Colorado, Connecticut, Hawai’i, Illinois, Indiana, Kentucky, Maine, Maryland, Minnesota, New Jersey, New York (at least up to the date of new legislation), North Carolina, Ohio, Oregon, South Dakota, Virginia, Washington, and West Virginia.  In short, they lacked “control” over State agencies.

However, in keeping with its focus on State law, the Ninth Circuit wrote:

We deny mandamus relief to the Pennsylvania AG because Pennsylvania law explicitly grants his office control over the documents of nonparty state agencies. See 71 Pa. Stat. and Cons. Stat. § 732-208 (West 2025) (“The Office of Attorney General shall have the right to access at all times to the books and papers of any Commonwealth agency necessary to carry out his duties under this act.”). Therefore, the Pennsylvania AG cannot show clear error in the discovery orders.

Id. at *3.  Finally, “because the district judge already granted these states relief from the magistrate judge’s order,” the Petitions of Delaware, Kansas, Louisiana, Nebraska, Rhode Island, South Carolina, and Wisconsin were moot. Id.

VARIOUS TESTS FOR “POSSESSION, CUSTODY, OR CONTROL”

I looked at one of several earlier phases of this dispute in Possession, Custody, or Control of Responsive Information by States Suing Meta (Sep. 12, 2024), citing In Re: Social Media Adolescent Addiction/Personal Injury Products Liability Litigation, 2024 WL 4125618 (N.D. Cal. 2024).  There, the court applied the “legal control” test of possession, custody, or control.[3]

Meta had argued that Rule 34 requests were the proper vehicle because it would otherwise have to serve 200 subpoenas under Rule 45. In short, it argued that the State AGs had control over State agencies.

There are three different standards for determining the meaning of “control.”  See Isn’t It Time for a Uniform National Standard on “Possession, Custody, or Control”? (Sept. 7, 2022)(citations omitted).  They are: 1) the “legal right” to control standard; 2) the “legal right plus notification” standard; and, 3) the “practical ability” to control standard.  Id.; The Sedona Conference, “Commentary on Rule 34 and Rule 45 ‘Possession, Custody, or Control,’” 17 Sedona Conf. J. 467, 482-92 (2016).

To complicate the matter, the Sedona publication states “that federal courts in some circuits have applied more than one standard.” Id. at 492. The choice of the test may, and often does, impact the outcome of a discovery dispute.  Id.  That issue – which test applies, unlike the “control” issue – is governed by federal law.

In Social Media Adolescent Addiction, binding federal authority compelled use of the “legal right” test.  Therefore, “control” existed where a litigant had the legal right to obtain documents on demand.  And, the Ninth Circuit applied State law to the second part of that analysis, i.e., what is “control”?

THE DISTRICT COURT’S DECISIONS

Many of the “control” factors considered by the district court, and discussed at length in Possession, Custody, or Control of Responsive Information by States Suing Meta, were not discussed in the Ninth Circuit’s opinion.  The Magistrate Judge’s opinion, 2024 WL 4125618, is 146 pages on Westlaw and cannot be summarized here.

Subsequently, the District Judge issued a 38-page opinion.  2025 WL 1179370 (N.D. Cal. Mar. 6, 2025)(describing the Magistrate Judge’s opinion as 248 pages).  With the exception of nine States, the District Judge found “the [Magistrate Judge’s] Order correct, well-reasoned, and thorough, and rejects most objections.”

The State AGs’ objections in the District Court related to the Magistrate Judge’s holding on “legal control.”  The State AGs made a “dual-executive” argument; however, the District Judge wrote: “Taking that as true—and the Court sees nothing objectionable in that characterization—the States miss the point. An entity can still have legal control over the documents of an entirely independent entity….”  Id. at *3.  It  wrote: “If the mere fact that two separately constituted or formed entities were enough to defeat control, then there would almost never be a finding of control.”  Id. at *4.

The District Court held that the “dual executive” theory neither permits nor defeats control. “The question is whether there is some mechanism that confers on a given AG the legal right to obtain documents of a state agency on demand….” Id.

Further, the District Court noted that, in 32 of the cases, the State itself was a party.  Both the District Judge and Magistrate Judge pointed to, among other things, the fact that the State AGs had sent litigation hold notices to the agencies, suggesting that was an indicator of control. Id. at *8.

The District Judge wrote: “It is particularly dilatory for these agencies to sit back and rely on their AG to effectively act as their counsel by advocating on their behalf over months of litigation, now to disclaim any such association with their AG.” Id.  The lower court also wrote: “Ultimately, the state agencies protesting discovery miss the point: whether by party or third-party discovery, if a document is relevant, it will be produced. In federal court, federal law dictates the process by which that occurs.”  Id. In summary:

Many states provide that the AG is charged to supervise the general legal affairs of the state and that state agencies may not obtain separate counsel without “the written consent of the Attorney General.” See, e.g., Cal. Gov’t Code § 11040(c). In other words, the AG has the effective authority to mandate that it represent the agency in connection with this litigation….  The [Magistrate Judge’s] Order identified that, as a matter of federal law, this mechanism provides an AG with the authority to control this litigation on behalf of state agencies and by extension have a right to access relevant documents in their possession for purposes of discovery. Under this deferential standard of review, the Court cannot say that such determination is contrary to law or clearly erroneous.

Id. at *9.  After an analysis of each State’s laws, the District Judge found clear error as to seven or, perhaps after supplementation, nine states.  Id. at *31. The District Court wrote:  “In federal court, federal law governs the scope of discovery. The States cannot employ the benefits of an MDL without the attendant burdens.” Id. at *31.

It is from that decision that California and others petitioned for mandamus.

WHILE “CONTROL” APPLIES TO DISCOVERY,

IT ALSO APPLIES TO PRESERVATION

The definition of “control” may be important in another context.  In addition to practical consequences in discovery – such as here, the need to serve perhaps hundreds of subpoenas – the definition of “control” may impact the duty to preserve.

Sedona Principle 5 provides for notice of lack of legal control by the responding party and adds: “If the responding party so notifies the requesting party, absent extraordinary circumstances, the responding party should not be sanctioned or otherwise held liable for the third party’s failure to preserve the Documents or ESI.” [emphasis added]. Applying that  principle, if California notified the Defendants that it does not control State agency documents for discovery, the California plaintiff cannot be sanctioned if a State agency destroyed relevant, discoverable information after the plaintiff’s duty to preserve was triggered.[4]

In the prior blog, I wrote that “for my 2¢,” the “practical ability” to control test is preferable to the “legal control” test.  Let me give a simple example.  A and B are married. Each has their own email account. If A were to ask B to provide a copy of an email, and B as a loving spouse would do it, A has “practical ability” and therefore control.

But, as a spouse, A does not have legal control over B or B’s email account, and, under that test, an opponent must subpoena B. Further, if A notifies opposing counsel of the lack of control, and if Sedona Principle 5 governs, prior to service of a subpoena on B, B can destroy the email (after A’s duty to preserve is triggered) and neither A nor B is subject to sanction.

Fed.R.Civ.P.1 calls for the just, speedy, and inexpensive resolution of every action.  I suggest that and other hypotheticals should be troubling to the “legal control” proponents.

However, in my prior blog, I added: “The debate over the varying standards that is reflected in the various judicial decisions and The Sedona Conference Commentary shows that there are strengths and weaknesses in each of the formulations.  If any one was perfect, the choice would be easy.  It isn’t.  I don’t have the answer to that question.”  There needs to be a uniform, national standard. Isn’t It Time for a Uniform National Standard on “Possession, Custody, or Control”?

THANK YOU

I would like to thank Ethan Ackerman, Esq., of Morgan,  Lewis & Bockius LLP, for bringing In Re: People of the State of California v. U.S. District Court for the Northern District of California, Oakland, 2025 WL 2427608 (9^th Cir. Aug. 22, 2025), to my attention.

Mr. Ackerman “counsels clients in complex litigation and all levels of governmental investigations, with a focus on issues of electronic discovery…. With an emphasis on proactive information governance, Ethan tailors plans for effectively handling and minimizing data, and offers prelitigation counseling regarding legal hold and document preservation, information management, and records retention policies.”

Of course, only I am responsible for the content of this blog.  The opinions and positions expressed on this site are mine alone.

____

[1] In a different context, it was improper to name the Court and Judges as defendants.  Lawsuit Against Federal Judges Dismissed – “[T]hese are not normal times.” (Aug. 26, 2025).

[2] Lockyer was apparently decided with the State as an enforcing plaintiff.  In Meta, the States were plaintiffs and there was no need to analyze “possession, custody, or control” where the State is a defendant.  Assume, for example, that a State department of transportation operates a subordinate transit police agency which is sued for excessive force in an arrest. Assume that the State AG represents the police agency and the non-party transportation department, which has relevant, discoverable documents.  In a defensive action, does the State AG have possession, custody, or control over the department of transportation documents?

[3] For blogs discussing another prior decision in Social Media Adolescent Addiction, see also   4 Things You Never Wanted to Hear + 8 Lessons & a Dispute Over Revealing Recipients of Litigation Hold Notices, All in One Case (May 4, 2024); An Epilog:  4 Things You Never Wanted to Hear + 8 Lessons & a Dispute Over Revealing Recipients of Litigation Hold Notices, All in One Case (May 10, 2024).  I concluded by writing: “This is most certainly not the last chapter in In Re: Social Media Adolescent Addiction.”

[4] In Silvestri v. GMC, 271 F.3d 583 (4th Cir.2001); Zorzit v. Comptroller, 225 Md. App. 158, 180, 123 A.3d 627, 640 (2015), the courts stated that, when a party cannot preserve potentially responsive information, it has a duty to notify the opponent.

Thompson I dealt in part with the December 2000 amendment to Fed.R.Civ.P. 26 which changed the scope of discovery from “any matter, not privileged, which is relevant to the subject matter involved in the pending action,” to unprivileged facts relevant to “the claim or defense of any party,” unless the court determines that there is “good cause” to permit broader discovery.[2]

The Thompson I Court wrote:

Although the rule changes do not specifically explain the difference in scope between discovery relevant to “claims and defenses” in the litigation, and discovery relevant to the “subject matter”, it is clear that the former is intended to be narrower than the later, and that the broader discovery is only to be allowed for “good cause”. Further, it seems clear that the most valuable reference to use in implementing the new change in the scope of discovery is the pleadings that have been filed, as that is where the claims and defenses are stated. However, the pleadings are only the starting place…. [Emphasis added].

Importantly, the Court noted that the dividing line between the two standards “cannot be defined with precision….” [citation omitted].

On the one hand, it wrote: “Thus, counsel should be forewarned against taking an overly rigid view of the narrowed scope of discovery. While the pleadings will be important, it would be a mistake to argue that no fact may be discovered unless it directly correlates with a factual allegation in the complaint or answer.”

On the other hand, the Thompson I Court noted: “It equally is clear, however, that the new rule represents a change from the old version, and that, unless expanded by the court for good cause shown,[3] it is intended that the scope of discovery be narrower than it was, in some meaningful way.”

The 2000 GAP Report to Rule 26(b)(1) explains that the change from “subject matter” discovery had been first proposed, and rejected, in 1978.  It wrote:

The Committee has heard that in some instances, particularly cases involving large quantities of discovery, parties seek to justify discovery requests that sweep far beyond the claims and defenses of the parties on the ground that they nevertheless have a bearing on the “subject matter” involved in the action….

The Committee intends that the parties and the court focus on the actual claims and defenses involved in the action….  The rule change signals to the court that it has the authority to confine discovery to the claims and defenses asserted in the pleadings, and signals to the parties that they have no entitlement to discovery to develop new claims or defenses that are not already identified in the pleadings. In general, it is hoped that reasonable lawyers can cooperate to manage discovery without the need for judicial intervention. When judicial intervention is invoked, the actual scope of discovery should be determined according to the reasonable needs of the action. The court may permit broader discovery in a particular case depending on the circumstances of the case, the nature of the claims and defenses, and the scope of the discovery requested.

The Thompson I Court offered the following analysis:

Lest litigants and the court become consumed with the philosophical exercise of debating the difference between discovery relevant to the “claims and defenses” as opposed to the “subject matter” of the pending action—the juridical equivalent to debating the number of angels that can dance on the head of a pin—the practical solution to implementing the new rule changes may be to focus more on whether the requested discovery makes sense in light of the Rule 26(b)(2) factors,[4] than to attempt to divine some bright line difference between the old and new rule. Under this approach, when confronted with a difficult scope of discovery dispute, the parties themselves should confer, and discuss the Rule 26(b)(2) factors, in an effort to reach an acceptable compromise, or narrow the scope of their disagreement.

The Court suggested:

This means that the parties must set aside their differences as adversaries and make a good faith effort to reach common ground on the disputes. It strikes me that this case is a perfect example of how creative counsel can employ the phasing methods used by the Court in Marens to permit the plaintiffs to have access to some, but less than all, of the information they seek, with the understanding that if, following the initial, limited review, additional discovery would make sense under the Rule 26(b)(2) factors, it will be provided. Cost shifting or sharing also should be considered.

In my opinion, Thompson I stands for the proposition that “claims and defenses” discovery is narrower than “subject matter” of the action discovery, and the former is best defined by a review of the pleadings that is guided by proportionality considerations.

_____

[1] For a discussion of Thompson II, please  see the first blog in this series.

[2] This blog does not address the December 2015 amendments and Fed.R.Civ.P. 26(b)(1) now reads: “Unless otherwise limited by court order, the scope of discovery is as follows: Parties may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense and proportional to the needs of the case, considering the importance of the issues at stake in the action, the amount in controversy, the parties’ relative access to relevant information, the parties’ resources, the importance of the discovery in resolving the issues, and whether the burden or expense of the proposed discovery outweighs its likely benefit. Information within this scope of discovery need not be admissible in evidence to be discoverable.”  The “good cause” expansion was deleted.

[3] Id.

[4] The reference to Rule 26(b)(2) pre-dates the 2015 amendments.  See Fed.R.Civ.P. 26(b)(1).