| Court - Judge Name |
Effective Date |
Applicable To | Categories |
Summary |
| N.D. Cal. – Magistrate Judge Nathanael M. Cousins | 11/20/2025 | Generative AI Usage | In Buchanan v. Vuori, Inc., Case No. 23-cv-01121-NC (N.D. Cal. Nov. 20, 2025), following the court’s order to show cause why plaintiff’s counsel should not be sanctioned for the inclusion of “false case quotations,” the attorney admitted to using “about six different AI tools to prepare his motion” (“ChatGPT-4, OpenAI, Claude and Claude Code, Clear Brief, Lexis Nexis, and Westlaw,” ECF 92 at 2), but admitted that “the AI tools failed to catch one another’s hallucinations.” The court opined that “[u]sing AI to check the work of AI was not a reasonable inquiry” into the facts and law, and that the attorney violated Federal Rule of Civil Procedure 11, Civil Local Rule 11-4, and California Rule of Professional Conduct 3.3. The court made clear that it was not his “use of AI that violated Rule 11, but rather his failure to conduct a reasonably inquiry into the law cited in his motion.” The court therefore imposed the following sanctions: striking plaintiff’s second motion for preliminary approval of a class action settlement and corrected motion without leave to refile; ordered plaintiff’s counsel to pay $250 to the clerk of court; referred counsel to the Court’s Standing Committee on Professional Conduct; and found the attorney inadequate to serve as class counsel. | |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| Cal. Ct. App. – Judge Frank J. Menetrez | 11/17/2025 | Generative AI | Generative AI Usage | In Schlichter v. Kennedy, 2025 Cal. App. LEXIS 740 (Nov. 17, 2025), the court found that sanctions were warranted for appellant’s counsel for “relying on fabricated legal authority” in his writ of supersedeas and opening brief, in violation of California Rules of Court 8.204(a)(1)(B). The rule provides that each brief must be supported “by argument and, if possible, by citation of authority.” Counsel’s opening brief and writ contained two and three citations, respectively, to cases that do not exist. The nonexistent cases shared the names of existing cases, but the other parts of the citation were incorrect, and the real cases did not support counsel’s arguments. Judge Menetrez ordered counsel to provide official copies of the fabricated cases. Counsel asserted that he meant to cite the existing cases: he provided copies of the real cases that matched the fabricated case names but failed to address the fact that the volume, page numbers, and legal propositions were different than those in the cases he cited. In response to the court’s subsequent order to show cause, counsel reasserted this position. He admitted to accidentally misciting the cases “at the reporter/volume/page level” but argued that the cases were real California cases, that the errors were merely “clerical in nature,” and rejected as “entirely untrue” the allegation that he used “‘AI hallucination’ cases.” At a hearing, however, counsel admitted that he regenerated the brief using GenAI to reorganize it but stated that he could not recall whether he used GenAI in preparing the writ (after initially denying using “any AI whatsoever” to generate the writ). However, he continued to maintain that the errors were not the result of GenAI use. The court did not find these assertions credible. Judge Menetrez noted that “[t]he spurious citations do not involve the mere omission or addition or transposition of one or several digits,” but were “completely different from the correct citations for the actually existing cases that have those case names.” Furthermore, the real cases supported legal propositions so different than those in counsel’s filings that “it would make no sense for [him] to claim that he intended to cite the actually existing cases to support those legal propositions.” Judge Menetrez also found that counsel lacked “candor and credibility” in other parts of his response. The court highlighted the importance of ensuring each citation is correct and honesty to the court, issuing the attorney $1,750 in sanctions. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| E.D. Cal. Magistrate Judge Chi Soo Kim | 10/14/2025 | Generative AI | Generative AI Usage | In Pyle v. Glenn Cnty. Sheriff's Dep't, 2025 U.S. Dist. LEXIS 202630 (E.D. Cal. Oct. 14, 2025), the pro se plaintiff filed a sur-reply that contained GenAI “prompts and responses to prompts.” The court reminded the pro se plaintiff that the “Federal Rule of Civil Procedure 11 applies to all parties, including parties such as Plaintiff who are proceeding without an attorney” and cautioned, that while the court has not prohibited the use of GenAI, “pleadings and filings that include ‘hallucinations’ from the use of generative AI (i.e., made up facts or legal authority) can violate Rule 11.” The court further cautioned that a violation of Rule 11 can result in sanctions. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal. – Judge Araceli Martínez-Olguín | 10/10/2025 | Generative AI | Generative AI Usage | In Oneto v. Watson, No. 22-cv-05206-AMO, Dkt. No. 94 (N.D. Cal. Oct. 10, 2025), plaintiff’s counsel included various hallucinated and erroneous case citations in his motion for judgment under Federal Rule of Civil Procedure 52. In response to the court’s order to show cause, the attorney admitted that he may have “unknowingly” copied case information “from an AI overview” on Google or another “open browser tab” that he intended to delete from the brief since he “had not read [those cases].” See id. at 4 (cleaned up). The court found the plaintiff’s attorney to be in violation of Federal Rule of Civil Procedure 11 for submitting a filing with inaccurate and unreviewed assertions and of the court’s Civil Standing Order for “fail[ing] to recognize the use of AI in the preparation of his papers.” Id. As a result, Judge Martínez-Olguín ordered (i) plaintiff’s counsel to pay a $1,000 penalty, serve a copy of this order on his client, and submit a declaration to the court that certifies his participation in a CLE course on the ethical use of AI; and (ii) the Clerk of the Court to serve a copy of this order on the State Bar of California. Id. at 7-8. |
| Applies to AI Used for Research | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| Court of Appeal of California – Justice Thomas A. Delaney |
9/15/2025 | Generative AI | Generative AI Usage | In Huynh v. Desimone, 2025 Cal. App. Unpub. LEXIS 5853 (Cal. Ct. App. Sept. 15, 2025), the pro se plaintiff-appellant included citations to “fictitious cases” in her brief on appeal of the family court’s order determining jurisdiction. The defendant-appellee, in support of their motion to dismiss the appeal, flagged these citations and requested that the court sanction the plaintiff-appellant. Justice Delaney ultimately declined to impose sanctions because the defendant-appellee failed to formally move the court to impose sanctions and instead included such request in their opposition brief and respondent’s brief. However, the court cautioned future pro se parties that they “may not be so fortunate to avoid monetary or other sanctions,” as the work associated with reviewing fictitious cases ‘damages the litigant’s credibility with the court” and “wastes the limited resources of the courts.” |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| Court of Appeal of California – Justice Lee Smalley Edmon |
9/12/2025 | Generative AI | Generative AI Usage | In Noland v. Land of the Free, L.P., 2025 Cal. App. LEXIS 584 (Cal. Ct. App. Sept. 12, 2025), “nearly all” of the citations in the plaintiff-appellant’s opening and reply briefs cited to non-existent cases or included quotes that could not be identified in those cases. During the hearing on the court’s order to show cause as to why the court should not sanction plaintiff’s counsel for “filing appellate briefs replete with fabricated quotes and citations,” counsel admitted that he “relied on AI ‘to support citation of legal issues’” and was unaware that GenAI tools could create hallucinated sources and citations. As part of the drafting process, plaintiff’s counsel submitted his first drafts of the briefs to ChatGPT for “enhance[ments]” but did not review the “enhanced” versions of the briefs before filing and also relied on Claude, Gemini, and Grok when preparing these briefs. Judge Edmon ultimately ordered (i) plaintiff’s counsel to pay a $10,000 penalty and send a copy of this opinion to his client and (ii) the clerk of the court to forward a copy of this opinion to the State Bar. In doing so, Judge Edmon reminded parties of the court’s authority to sanction parties for filing frivolous appeals under Section 907 of the Code of Civil Procedure and for “unreasonably” violating the California Rules of Court, specifically those that require assertions to be supported “if possible, by citation of authority.” Judge Edmon also reiterated that her opinion also serves as a warning to all California state court litigants that “no brief, pleading, motion, or any other paper filed in any court should contain any citations—whether provided by generative AI or any other source—that the attorney responsible for submitting the pleading has not personally read and verified.” Attorneys must also “read the legal authorities they cite in appellate briefs or any other court filings.” If plaintiff’s counsel had “read the cases he cited before filing,” he would have discovered the various citation errors and would not have “abdicated his responsibility to the court and to his client.” |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| C.D. Cal. – Judge Percy Anderson |
9/12/2025 | Generative AI | Generative AI Usage | In Yi-Sheng Fang v. Hechalou, 2025 U.S. Dist. LEXIS 181749 (C.D. Cal. Sept. 12, 2025), the defendants submitted an errata sheet to address their inclusion of four non-existent quotations and two incorrect case citations in their opposition to plaintiffs’ motion for a preliminary injunction. However, in response to the errata sheet, the plaintiffs submitted a supplemental declaration that spotted three additional hallucinated citations. The defendants then filed a supplemental declaration in response, noting that, “as a small firm with ‘fee sensitive’ clients,” the firm uses AI tools to “more effectively undertake cases and provide access to justice,” and the attorney responsible for drafting the opposition violated firm policy in “fail[ing] to review and confirm” the accuracy of all cited material. Since this incident, the firm noted that the attorneys who worked on the opposition have since agreed to take CLE courses on “the use of AI in legal writing,” and the filing attorney expressed his apologies to the court and admitted that he failed to confirm that all cited authority “existed or represented the legal positions asserted.” Though the court appreciated the attorney’s apology, Judge Anderson found these mistakes to be “highly problematic and simply unacceptable” in light of the numerous “recent and highly publicized cases” discussing GenAI misuse. The court ultimately ordered defendants’ counsel’s law firm to pay 50% of plaintiff’s attorneys’ fees associated with the review of the hallucinated citations and did not sanction the individual attorneys after a consideration of the errata sheet and other remediating efforts. In addition, defendants’ counsel was ordered to send a copy of this order to the California State Bar and file a declaration of compliance associated with this order. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms – Attorneys/Law Firms | ||||
| Cal. Super. Ct. -- Judge Victoria Kolakowski | 9/9/2025 | Generative AI | Generative AI Usage | In Mendones, et al. vs. Cushman and Wakefield, Inc., et al., No.: 23CV028772 (Cal. Super. Ct. Sept. 9, 2025), the court issued a terminating sanction (i.e., striking the second amended complaint and dismissing the action with prejudice) against plaintiffs for intentionally submitting false evidence involving GenAI deepfake video evidence to the Court in connection with their motion for summary judgment. After reviewing the video evidence submitted by plaintiffs, the Court assessed that the “lack of facial expressions, the looping video feed, among other things, suggested that these exhibits were products of GenAI—i.e., ‘deepfakes.’” The Court issued a tentative ruling, including an order for plaintiffs to show cause why sanctions were not warranted. The plaintiffs submitted declarations, but the court found the plaintiffs’ explanation for the false evidence lacking. Additionally, the plaintiffs failed to appear at an in-person show cause hearing. The Court found that while the GenAI misuses warranted terminating sanctions, it did not warrant monetary sanctions because more severe sanctions were warranted. Lastly, the court contemplated a referral for criminal prosecution for perjury and forgery, but found that it was too severe and would not address the harm caused by the plaintiffs in the case. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| E.D. Cal. – Judge Dena Coggins |
9/8/2025 | Generative AI | Generative AI Usage | In Tercero v. Sacramento Logistics, LLC, 2025 U.S. Dist. LEXIS 175962 (E.D. Cal. Sept. 8, 2025), the defendants’ opposition to the plaintiff’s motion for reconsideration flagged 24 cases the plaintiff cited that were either nonexistent, did not contain the purported quotations in the brief, or did not support the plaintiff’s propositions in the brief. In response, the plaintiff’s attorney stated that some citations contained no errors or, to the extent they did, they amounted to small pincite or case name errors. The plaintiff’s attorney also submitted a declaration that contained allegedly corrected citations and requested that the court strike the fourteen citations with errors. During the hearing on the court’s order to show cause, the plaintiff’s attorney denied any use of artificial intelligence and claimed her errors “stemmed from inaccuracies in [her] handwritten notes.” However, upon closer review, the court found that the proposed new citations still did not rectify plaintiff’s errors. The inclusion of these citations violated both Local Rule 180(e) for failing to comply with the State Bar of California’s standards of professional conduct and Rule 11 of the Federal Rules of Civil Procedure for making unsupported assertions and false attributions of authority. As such, Judge Coggins imposed a $1,500 penalty on plaintiff’s counsel and ordered the Clerk of the Court to serve a copy of this order on the State Bar of California. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms – Attorneys/Law Firms | ||||
| California Rules of Court | 9/1/2025 | Generative AI | Generative AI Usage | Standard 10.80 of the California Rules of Court entitled “Use of generative artificial intelligence by judicial officers” provides guidance to “judicial officers” (i.e., judges) on permissive and prohibited GenAI use in the court. Notably, state judges are advised to “take reasonable steps” to (i) ensure that any GenAI-generated material either created themselves or by a third party is truthful and accurate and (ii) rectify any incorrect GenAI-generated material. Judges are also to remain aware of any personal identifying information or discriminatory/biased content in GenAI tools and consider whether to publicly disclose the use of GenAI in public-facing materials. |
| Requires Disclosure and/or Verification | ||||
| C.D. Cal. – Judge Dolly M. Gee | 8/12/2025 | Generative AI | Generative AI Usage | In Twist IT Up, Inc. v. Annie Int’l, Inc., 2025 U.S. Dist. LEXIS 170025 (C.D. Cal. Aug. 12, 2025), the plaintiff included various allegedly hallucinated citations in its opposition to the defendant’s motion to strike and failed to remediate the erroneous citations. Following an order to show cause, the court ultimately imposed a $500 fine on the plaintiff’s attorney for submitting a brief with “invalid case citations” and required the attorney to serve the hearing minutes on his client. ECF No. 44. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms – Attorneys/Law Firms | ||||
| C.D. Cal. – Judge Anne Hwang | 7/31/2025 | Generative AI | Generative AI Usage | In Martinez v. Dollar Tree Stores, Inc., 2025 U.S. Dist. LEXIS 147616 (C.D. Cal. July 31, 2025), the pro se plaintiff included a hallucinated case in her motion for remand, and the court held that such conduct failed to comply with the court’s standing order that requires parties who use GenAI to disclose such use and certify that all cited authority is accurate. Although the court did not sanction the plaintiff for her GenAI misuse and her failure to comply with the standing order, Judge Hwang cautioned that future citations to hallucinated authority will result in a show cause order where the plaintiff would have to explain why she should not be sanctioned. |
| Requires Disclosure and/or Verification | ||||
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| C.D. Cal. – Judge Anne Hwang | 7/31/2025 | Generative AI | Generative AI Usage | In Mirage v. Costco Wholesale Corp., 2025 U.S. Dist. LEXIS 147615 (C.D. Cal. July 31, 2025), the pro se plaintiff included a hallucinated case in her motion for remand, and the court held that such conduct failed to comply with the court’s standing order that requires parties who use GenAI to disclose such use and certify that all cited authority is accurate. Although the court did not sanction the plaintiff for her GenAI misuse and her failure to comply with the standing order, Judge Hwang cautioned that future citations to hallucinated authority will result in a show cause order, where the plaintiff would have to explain why she should not be sanctioned. |
| Requires Disclosure and/or Verification | ||||
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal. – Judge Edward M. Chen | 7/30/2025 | Generative AI | Generative AI Usage | In Coronavirus Rep. Corp. v. Apple Inc., 2025 U.S. Dist. LEXIS 146572 (N.D. Cal. July 30, 2025), the plaintiffs used ChatGPT to draft an exhibit calling for Tim Cook’s resignation submitted in connection with three different motions and filed a separate motion with AI-generated citations. Plaintiffs’ counsel admitted that “no human ever authored the Petition for Tim Cook’s resignation,” and that he submitted a specific prompt to ChatGPT asking whether “Judge Gonzales Rogers’ rebuke of Tim Cook’s Epic conduct create[d] a legally grounded impetus for his termination as CEO, and if so, write a petition explaining such basis, providing contextual background on critics’ views of Apple’s demise since Steve Jobs’ death.” Plaintiffs’ counsel further conceded that the entire exhibit was submitted as a “test” and “was reviewed in under ‘fifteen minutes.’” In addition to this ChatGPT output, the plaintiffs' anti-SLAPP motion and response to the court's order to show cause contained various hallucinated citations and citations to inappropriate cases. As sanctions for the hallucinated citations and other conduct, the defendants sought (i) $400,000 in joint and several compensatory penalties; (ii) the revocation of plaintiff's counsel's pro hac vice admission; and (iii) a pre-filing injunction. Judge Chen ultimately ordered the defendant to submit time sheets for their work performed "as a result of Plaintiffs' sanctionable conduct" but denied the request for the revocation of the pro hac vice status and pre-filing injunction, as "an attorney appearing on behalf of a client cannot be sanctioned as a vexatious litigant." |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms – Attorneys/Law Firms | ||||
| Court-Imposed Consequences – Attorneys/Law Firms – Parties | ||||
| N.D. Cal. – Judge Donna M. Ryu | 7/25/2025 | Generative AI | Generative AI Usage | In Pop Top Corp. v. Rakuten Kobo Inc., 2025 U.S. Dist. LEXIS 143212 (N.D. Cal. July 25, 2025), the pro se judgment debtor/interested party included hallucinated cases in both his motion for admonishment regarding litigation and procedural abuse and his notice of correction regarding case citations. Although the pro se litigant claimed that the citations to these cases were inadvertently included, he provided no reason as to how they ended up in the briefing and “repeatedly tried to invoke his self-represented status as a shield.” Judge Ryu was sympathetic to his pro se status but reminded him that all parties—notwithstanding their status of representation—must ensure that all assertions are grounded in existing authority. The court ultimately declined to impose sanctions at this time but warned that further use of hallucinated citations or quotations could result in sanctions. |
| Requires Disclosure and/or Verification | ||||
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| Judicial Council of California – Artificial Intelligence Task Force | 7/18/2025 | Generative AI | Generative AI Usage | The Artificial Intelligence Task Force of the Judicial Council of California recommended a rule ordering all California state courts that do not currently prohibit court staff’s or judicial officers’ use of GenAI to adopt a GenAI use policy by December 15, 2025. |
| C.D. Cal. – Judge Fernando M. Olguin | 7/1/2015 | Any AI | Any AI Usage | Under Section IV(A)(5) of his standing order, Judge Olguin requires that “[i]f any party or attorney uses an artificial intelligence tool in the preparation of any filing, the submission of that document signifies that the individual responsible for the filing has certified that she/he reviewed all source material and verified the accuracy of any AI content.” Judge Olguin also cites to Federal Rule of Civil Procedure 11, indicating some possibility that noncompliance with his rule may implicate the federal rule governing sanctions. Although Judge Olguin’s standing order does not require additional steps to reflect or disclose the use of AI, his order makes clear that the party or attorney’s submission of the document itself constitutes an affirmation that it has been reviewed for accuracy. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Research | ||||
| Applies to AI Used for Filings/Drafting | ||||
| E.D. Cal. – Judge Daniel J. Calabretta | 6/23/2025 | Generative AI | Generative AI Usage | In Allen v. Mercedes-Benz USA, LLC, 2025 WL 1744451 (E.D. Cal. June 24, 2025), the plaintiff’s counsel included two “fabricated quotations” in a reply brief. When ordered to illustrate the source of the quotations, counsel neglected to do so and included another false citation in his response. Judge Calabretta found that plaintiff’s counsel violated both Rule 11 of the Federal Rules of Civil Procedure for failing to “investigate the caselaw” and Local Rule 180(e) for failing to adhere to the California bar’s ethics rules against “knowingly mak[ing] a false statement of fact or law . . . or fail[ing] to correct a false statement of material fact or law.” The court referred the matter to the State Bar of California. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| Cal. Super. Ct. -- Judge Frederick S. Chung | 5/29/2025 | Generative AI | Generative AI Usage | In Krishnakumar v. Eichler Swim and Tennis Club, Case No. 24-CV-432638 (Cal. Super. Ct. May 29, 2025), the court found that the defendant’s “principal citation . . . appears to be completely made up” after spending “an inordinate amount of time searching for this case.” The court theorized that it was either “generated by an AI tool” or was an “intentional deception.” Although the court concluded that the latter scenario is both more concerning and more likely given it was the defendant’s “key legal citation”—noting that such a deception would be a “violation of Rule 3.3 of the California Rules of Professional Conduct”—the court did not impose penalties for the fake case, but ruled against the defendant on the merits. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal. -- Magistrate Judge Susan Van Keulen | 5/23/2025 | Generative AI | Generative AI Usage | In Concord Music Group, Inc. v. Anthropic PBC, 2025 WL 1482734, the defendant’s expert declaration included a hallucinated citation to a non-existent internet article. Although the defendant initially painted this as an “honest citation mistake,” Magistrate Judge Van Keulen found this to be a “serious” issue, especially because the defendant allegedly conducted a manual citation review and still failed to catch this error. The court also found the submitted declaration to be in violation of District Judge Lee’s Civil Standing Order that requires lead counsel to personally verify the accuracy of all filings. As a result, the court struck the expert declaration in part and noted that the hallucinated citation “undermine[d]” the declaration’s overall credibility. |
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| Court-Imposed Consequences – Parties | ||||
| C.D. Cal - Judge Anne Hwang | 5/20/2025 | Generative AI | Generative AI Usage | Judge Hwang’s standing order requires any filings containing content created by GenAI to include a separate declaration “disclosing the use of [generative] artificial intelligence and certifying that the filer has reviewed the source material and verified that the artificially generated content is accurate and complies with the filer’s Rule 11 obligations.” |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal. -- Judge Edward M. Chen | 5/19/2025 | Generative AI | Generative AI Usage | In Gjovik v. Apple Inc., 2025 U.S. Dist. LEXIS 95303, the defendant requested that the court impose sanctions on the pro se plaintiff under Rule 11 of the Federal Rules of Civil Procedure for using GenAI when drafting various filings during the course of the litigation. According to the defendant, the plaintiff has yet to verify that the GenAI-generated information used in her filings was accurate. The court rejected the defendant’s request to impose sanctions but cautioned the plaintiff that she is ultimately responsible for the accuracy of information included in briefing before they are filed on the docket. The court further warned that a continued failure to verify this information could lead to a finding of contempt and the plaintiff’s inability to proceed pro se. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| C.D. Cal. -- Special Master Michael R. Wilner | 5/5/2025 | Generative AI | Generative AI Usage | In Lacey v. State Farm Gen. Ins. Co., 2025 U.S. Dist. LEXIS 90370, 2025 LX 26392, a case submitted to JAMS for alternative dispute resolution, the plaintiff’s “supplemental brief contained numerous false, inaccurate, and misleading legal citations and quotations.” Altogether, “nine of the 27 legal citations in the ten-page brief were incorrect in some way,” two of the cited authorities “do not exist at all,” and several quotations “were phony and did not accurately represent” the cited materials. Plaintiff’s attorneys admitted that one of the attorneys used a GenAI tool to draft an outline for the brief and that no attorney at either of the law firms representing the plaintiff reviewed the brief for accuracy. The court-appointed Special Master chastised the lawyers for their “bad faith” conduct through their “undisclosed use of AI,” their “failure to cite-check” the original brief, and “perhaps most egregiously . . . the re-submission of the defective Revised Brief without adequate disclosure of the use of AI,” describing such conduct as “reckless” and with the “improper purpose of trying to influence [the Special Master’s] analysis.” The Special Master concluded by saying “Directly put, Plaintiff’s use of AI affirmatively misled me.” As a consequence, the Special Master declined to consider the plaintiff’s supplemental briefs, declined to award any of the requested relief, and required that the law firms representing the plaintiff pay the defendant $26,100 as reimbursement for JAMS-related fees that had been previously paid by the defense, and required those firms to pay the defense an additional $5,000 for costs relating to defense counsel’s fees.
|
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| N.D. Cal. – Magistrate Judge Susan Van Keulen | 4/22/2025 | Generative AI | Generative AI | As part of the order granting in part and denying in part the defendant’s motion to dismiss in Yelp Inc. v. Google LLC, Magistrate Judge Van Keulen discussed one of the plaintiff’s allegations that was supposedly “generated by a query to Google’s Gemini AI tool.” Although the defendant’s “concerns about ‘pleading-by-bot’ [were] legitimate,” the court noted that parties are permitted to use AI under the both Federal Rules of Civil Procedure and the Northern District of California’s local rules and to file pleadings that lack evidentiary support, as long as they are consistent with Fed. R. Civ. P. 11. The court did not address Rule 11’s further implications, however, because the defendant did not argue that the plaintiff’s use of AI was unreasonable, nor did the defendant move for sanctions. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| Alameda County Superior Court – Judge S. Raj Chatterjee | 3/10/2025 | Generative AI | Generative AI | In Michael Evans, et al v. Execushield Inc., et al, Judge Chatterjee ruled that attorneys presenting fake citations to the court were not competent to represent the class of plaintiffs and would not be awarded fees if the plaintiffs prevail. Although the court declined to impose monetary sanctions, it did require the attorneys to file a copy of the court’s order “in all other actions currently pending in the Alameda County Superior Court in any department in which any one of Plaintiffs’ counsel is an attorney of record,” and to review all of their filings in the Alameda County Superior Court “in any currently pending action in which Plaintiffs’ counsel used the same ‘tool’ that resulted in the misrepresentations here and to file corrected pleadings if necessary.” |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| E.D. Cal. – Magistrate Judge Chi Soo Kim | 1/17/2025 | Generative AI | Generative AI Usage | In United States v. Hayes, 2025 WL 235531 (E.D. Cal. Jan. 17, 2025), after the defendant included fictitious citations in his motion to unseal a motion to compel, he subsequently filed a notice of errata to correct the “misstatements, erroneous citations, and unsupported quotations.” The court found that the main case upon which defense counsel relied had “all the markings of a hallucinated case created by generative artificial intelligence.” Although the attorney certified that he “did not use and has never used AI” to draft motions, he did not indicate where he found the case or how the citation was created. The court concluded that despite his efforts to “clarify the record,” he made “knowing and willful misrepresentations with the intent to mislead the Court” in violation of standards of professional conduct, particularly given his nearly “six years of experience.” Defense counsel was ordered to pay a $1,500 penalty, and the court ordered the Clerk of Court to serve a copy of its sanctions order on the other Bars of which the defendant’s counsel is a member and on all of the district and magistrate judges in the Eastern District of California to “deter[] the larger bar from repeating similar conduct.” The court ensured that none of these penalties would impact the defense counsel’s client. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| Orange County Superior Court – Judge Knill | 1/25/2024 | Generative AI | Generative AI Usage | Judge Knill’s standing order requires attorneys and self-represented parties, who have used generative AI “in the preparation of any complaint, answer, motion, brief, or other paper filed with the Court” to disclose “in a clear and plain factual statement . . . that AI has been used in any way in the preparation of the filing and [certify] each and every citation to the law, or the record in the paper, has been verified as accurate.” The order specifically mentions ChatGPT and Google Bard. |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| C.D. Cal - Judge Anne Hwang | 12/17/2024 | Generative AI | Generative AI Usage | Judge Hwang’s standing order requires any filings containing content created by Gen AI to include a separate declaration “disclosing the use of [generative] artificial intelligence and certifying that the filer has reviewed the source material and verified that the artificially generated content is accurate and complies with the filer’s Rule 11 obligations.” |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| C.D. Cal. – Judge Percy Anderson | 12/12/2024 | Generative AI | Generative AI Usage | Judge Anderson recently issued a decision in Mojtabavi v. Blinken, 2024 U.S. Dist. LEXIS 225418 (Dec. 12, 2024), sanctioning a pro se plaintiff for repeated “use[] [of] a text-generative artificial intelligence tool (e.g., ChatGPT)” to generate “fake case citations” is “unacceptable,” constitutes a “violation of Local Rule 11-3.9,” and “justifies dismissal of this action.” Although Local Rule 11-3.9 does not address artificial intelligence directly, it does concern the proper format for case citations. This decision indicates the court’s willingness to utilize existing local rules (which do not directly address AI) to follow the practice of some other courts in sanctioning parties’ use of GenAI to create “falsified or inaccurate case citations.” |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| Court-Imposed Consequences – Attorneys/Law Firms | ||||
| Court-Imposed Consequences – Parties | ||||
| C.D. Cal. – Judge Slaughter | 10/4/2024 | Generative AI | Generative AI Usage | Judge Slaughter’s standing order requires any filings containing content created by Gen AI to include a separate declaration “disclosing the use of [generative] artificial intelligence and certifying that the filer has reviewed the source material and verified that the artificially generated content is accurate and complies with the filer’s Rule 11 obligations.” |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| S.D. Cal. – Judge Robinson | 10/4/2024 | Generative AI | Generative AI Usage | Judge Robinson’s standing order reminds counsel that Rule 11, and other applicable standards of practice, requires due care in reviewing and filing work product created with the assistance of Gen AI tools. The order specifically states that AI tools are not prohibited, but also reminds counsel that violations of Rule 11 may result in sanctions or other corrective or disciplinary action. The order cites various case law to show: (i) that using Gen AI tools may result in misrepresentations and fabrications; (ii) the duties imposed by Rule 11; and, (iii) an example of an attorney who was referred to a grievance panel after a filing a brief with nonexistent cases. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. CA – Judge Lee | 8/16/2024 | Generative AI | Generative AI Usage | The order requires any filings containing content created by “ChatGPT or other such generative [AI]” must include a certification that lead trial counsel has personally verified the content’s accuracy, and that counsel must personally confirm the “accuracy of any research” conducted using generative AI tools. The order warns that failure to comply with this requirement is sanctionable. Notably, the order also requires counsel to maintain records of “all prompts or inquiries submitted to any generative AI tools” in the event those records become relevant at any point. |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Research | ||||
| Applies to AI Used for Filings/Drafting | ||||
| C.D. Cal. - Judge Oliver | Current as of 5/9/2024 | Generative AI | Generative AI Usage | Judge Oliver’s standing orders require parties using gen AI to “generate any portion of a brief, pleading, or other filing” to submit a separate declaration disclosing the use of gen AI and certifying the review and accuracy of any AI-generated content. This language appears to be specific to the use of gen AI in drafting court filings. |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| C.D. Cal. – Judge Blumenfeld | 3/1/2024 | Generative AI | Generative AI Usage | Judge Blumenfeld has updated his standing orders for civil cases to require parties using gen AI to “generate any portion of” a court filing, to submit a separately filed document disclosing the use of gen AI and certifying the review and accuracy of any AI-generated content. This language appears to be specific to the use of gen AI in drafting court filings, and does not appear to apply to the use of gen AI for research or other supportive, non-drafting tasks. |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal. - Judge Araceli Martínez-Olguín | 11/22/2023 | Generative AI | Generative AI Usage | This standing order requires lawyers to confirm the accuracy of content generated by “ChatGPT or other such tools,” and any submissions “containing AI-generated content” must include certification that lead trial counsel has personally verified the content’s accuracy. The order also holds counsel responsible for “maintaining records of all prompts or inquiries submitted to any generative AI tools” in the event that those records become relevant. |
| Requires Disclosure and/or Verification | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal - Judge Rita F. Lin | 11/9/2023 | Generative AI | Generative AI Usage | Judge Lin's standing order requires lawyers to "personally confirm for themselves" the "accuracy of any research" conducted by gen AI, and reminds them that "counsel alone bears ethical responsibility for all statements made in filings." The order explicitly does "not prohibit[]" use of gen AI tools, nor does it impose any disclosure or certification requirements upon parties. |
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Research | ||||
| Applies to AI Used for Filings/Drafting | ||||
| N.D. Cal - Magistrate Judge Peter H. Kang | 7/14/2023 | Generative AI | Generative AI Usage | Judge Kang’s standing order devotes nearly four pages to the topic of AI, starting by noting that gen AI is “distinguishable from other categories of AI,” and as such, although the order tends to refer to “AI,” it explicitly does not apply to more commonplace tools which happen to incorporate “AI” features, including legal research and spellchecking tools, but is instead focused on gen AI. Judge Kang’s order first reminds all litigants to comply with state, federal, and ABA rules and guidelines on the use of AI. Next, he requires disclosure of when documents filed with the court are “created or drafted” using AI (including requirements that counsel maintain records sufficient to identify which portions were drafted by AI), and prohibits the use of “AI-hallucinated” or “fictitious” citations, noting the possibility of sanctions in such cases. Judge Kang’s order also requires disclosure of “AI-generated evidentiary material.” Finally, the order requires that any parties using gen AI “keep[] records of all prompts or inquiries submitted to any [] third party AI tools” because of the potential confidentiality and privilege issues that may arise when using AI tools “own[ed] and operate[d]” by “[t]hird parties and non-parties.” |
| Requires Disclosure and/or Verification | ||||
| Suggests Cautious Use of AI | ||||
| Applies to AI Used for Filings/Drafting |
California
Jump to:
