Ledoux v. Outliers, Inc, No. 3:24-cv-05808-TMC (Western District of Wash. (W.D. WASH) 2026)

Case details
Full caption
JOANN LEDOUX v. OUTLIERS, INC
Country
United States
Jurisdiction
Federal
Court
Western District of Washington (W.D. WASH)
Decided
2026
ORDER ON SANCTIONS - 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 UNITED STATES DISTRICT COURT WESTERN DISTRICT OF WASHINGTON AT TACOMA JOANN LEDOUX, Plaintiff, v. OUTLIERS, INC. (d/b/a THESIS, THESIS NOOTROPICS, FIND MY FORMULA, and FORMULA), et al., Defendants. Case No. 3:24-cv-05808-TMC ORDER ON SANCTIONS I. INTRODUCTION In November and December 2025, Plaintiff Joann LeDoux submitted multiple filings containing dozens of inaccurate factual and legal citations which appeared “hallucinated” by a generative artificial intelligence tool. On February 4, 2026, the Court ordered Plaintiff’s counsel, Ms. Jocelyn Stewart, to show cause and explain why she should not be sanctioned for these errors. Dkt. 209 at 7. Ms. Stewart filed a sworn declaration on March 4, explaining that she had used two artificial intelligence programs—ChatGPT and Claude—to generate the false citations, that she was unaware of any falsities when she submitted the documents, and that she “accept[s] full Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 1 of 11
ORDER ON SANCTIONS - 2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 responsibility for the errors” and has “changed [her] practices to ensure this does not happen again.” Dkt. 245 at 10–25, 33. For the reasons discussed below, the Court ORDERS that Ms. Stewart be SANCTIONED pursuant to Federal Rule of Procedure 11. II. BACKGROUND AND PROCEDURAL HISTORY On December 2, 2025, the Court held oral argument on Defendants’ motions for partial summary judgment.1 Dkt. 184. During that hearing, Defendants raised concerns that Plaintiff’s recently-filed Daubert motion was “littered” with citations Defendants believed to be false and hallucinated by artificial intelligence. Dkt. 187 at 48 (citing Dkt. 176). Over the following weeks, Defendants identified multiple other seemingly erroneous citations in Plaintiff’s proposed amended complaint (Dkt. 173-1), Daubert motion (Dkt. 176), motion to strike affirmative defenses (Dkt. 178), expert evidence (Dkt. 177 at 288–315; Dkt. 194-13; Dkt. 194-15), and response to Defendants’ motion to strike experts (Dkt. 197). Dkt. 191 at 2–3; Dkt. 194 at 14–15; Dkt. 194-17; Dkt. 201 at 1–2. On February 4, 2026, the Court found that Ms. Stewart “submitted dozens of inaccurate factual and legal citations across at least five different filings.” Dkt. 209 at 6. These errors included citations to nonexistent sources, as well as fictitious quotations from real sources. Concerningly, two of Plaintiff’s expert reports contained citations to the same nonexistent or misattributed academic articles. Id. at 3–5. Although Ms. Stewart had attempted to correct hallucinated citations in her Daubert motion, the Court found that “[h]er explanations and corrections [were] incomplete at best and misleading at worst.” Id. at 5. For example, “Plaintiff’s ‘corrected’ Daubert motion replaced some false quotations with real ones but, in doing so, 1 The Court discussed the facts of this case in detail when it addressed the motions for partial summary judgment and will not do so again here. Dkt. 202 at 2–8. Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 2 of 11
ORDER ON SANCTIONS - 3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 completely changed their meaning.” Id. at 5–6 (citing Dkt. 192 at 2–3). The Court ordered Ms. Stewart to show cause and explain why she should not be sanctioned pursuant to Rule 11. Id. at 7. Specifically, the Court required Ms. Stewart to submit a sworn declaration and: (1) detail all incorrect, misleading, or nonexistent citations and all incorrect or misleading assertions submitted to this Court; (2) provide a thorough explanation for how the incorrect citations discussed in this order were found or generated; and (3) to the extent that she used any artificial intelligence programs in preparing the submissions discussed in this order or any other submissions to this Court, Plaintiff’s counsel shall include the names of those programs in her declaration. Id. Ms. Stewart filed a declaration in response to the order on March 4, 2026. Dkt. 245. She avers that she used Claude and ChatGPT to draft her submissions in this case, that these programs generated the false citations identified in the Court’s order, and that she “did not verify the output against the source before filing.” Id. at 10, 28–30. She attributes her mistake to an errant belief that the programs would correctly reference documents she provided in a “closed system” and that “reliable inputs would produce reliable outputs.” Id. at 5–6, 29. She “do[es] not contend that [she] met [her] obligations under Rule 11(b)” as she “did not conduct a reasonable inquiry into the accuracy of every citation and quotation before signing and submitting the filings.” Id. at 28. She also states that she has “fundamentally changed [her] practices regarding the use of automation/AI tools in legal work” and “no longer use[s] any AI tool to generate quotations or citations for filings submitted to any court.” Id. at 30. Ms. Stewart ends her statement with the following: I accept full responsibility for the errors described in this declaration. I have changed my practices to ensure this does not happen again. I am committed to maintaining the integrity of submissions to this Court and to every court in which I practice. I am deeply embarrassed and can only undertake each new filing as the chance to try to reclaim the trust and confidence of this Court and opposing counsel. I respectfully submit this declaration and ask the Court to consider the corrective measures I have already taken and am prepared to undertake in determining an appropriate sanction. Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 3 of 11
ORDER ON SANCTIONS - 4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Id. at 33. Although Defendants initially raised the false citations with this Court, they have not moved for sanctions. The Court therefore addresses Ms. Stewart’s Rule 11 violations sua sponte. III. LEGAL STANDARDS Federal Rule of Civil Procedure 11 “imposes a duty on attorneys to certify that they have conducted a reasonable inquiry and have determined that any papers filed with the court are well grounded in fact, legally tenable, and not interposed for any improper purpose.” Cooter & Gell v. Hartmarx Corp., 496 U.S. 384, 393 (1990) (internal quotation marks omitted). Rule 11 provides, in relevant part, that: By presenting to the court a pleading, written motion, or other paper—whether by signing, filing, submitting, or later advocating it—an attorney or unrepresented party certifies that to the best of the person’s knowledge, information, and belief, formed after an inquiry reasonable under the circumstances: . . . (2) the claims, defenses, and other legal contentions are warranted by existing law or by a nonfrivolous argument for extending, modifying, or reversing existing law or for establishing new law. Fed. R. Civ. P. 11(b). “On its own, the court may order an attorney, law firm, or party to show cause why conduct specifically described in the order has not violated Rule 11(b).” Fed. R. Civ. P. 11(c)(3). Courts have “wide discretion in determining the appropriate sanction for a Rule 11 violation.” Hudson v. Moore Bus. Forms, Inc., 836 F.2d 1156, 1163 (9th Cir. 1987). “A sanction imposed under this rule must be limited to what suffices to deter repetition of the conduct or comparable conduct by others similarly situated.” Fed. R. Civ. P. 11(c)(4). Additionally, sua sponte sanctions under Rule 11 “will ordinarily be imposed only in situations that are akin to a contempt of court.” United Nat’l Ins. Co. v. R&D Latex Corp., 242 F.3d 1102, 1116 (9th Cir. Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 4 of 11
ORDER ON SANCTIONS - 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 2001) (emphasis omitted) (quoting Barber v. Miller, 146 F.3d 707, 711 (9th Cir. 1998)). “[I]n the Ninth Circuit, the ‘akin to a contempt’ standard ‘generally requires the court to find bad faith or conduct tantamount to bad faith.’” Mavy v. Comm’r of Soc. Sec. Admin., No. CV-25-00689-PHX-KML (ASB), 2026 WL 91483, at *6 (D. Ariz. Jan. 13, 2026) (quoting Rocha v. Fiedler, No. 24-3692, 2025 WL 1219007, at *1 (9th Cir. Apr. 28, 2025)). IV. DISCUSSION Ms. Stewart concedes that she violated Rule 11 by failing to properly review her submissions to this Court, resulting in dozens of false citations. Dkt. 245 at 28, 33. The Court must therefore determine (1) whether her conduct was akin to contempt and (2) if so, what sanctions are appropriate. A. Ms. Stewart’s conduct was akin to contempt, and the Court may issue sanctions sua sponte. Ms. Stewart’s conduct was tantamount to bad faith and therefore akin to contempt. Her false citations “constitute fundamental, material, legal and factual errors submitted to the Court because AI-generated work product . . . was subjected to no discernable scrutiny.” Parker v. Costco Wholesale Corp., No. C25-0519-SKV, 2025 WL 4228413, at *11 (W.D. Wash. Nov. 7, 2025). “That failure was compounded by [Ms. Stewart] certifying, under Rule 11, that [s]he had conducted [a] reasonable inquiry into the supportability of h[er] filings knowing full well [s]he had not done so.” Id. Ms. Stewart complied with the Court’s order to show cause—admitting that she generated the false citations, accepting full responsibility for her conduct, and noting that she had ceased using artificial intelligence “to generate quotations or citations for filings submitted to any court.” Dkt. 245 at 28–30, 33. While the Court appreciates Ms. Stewart’s responsiveness, several factors weigh toward a finding of bad faith. Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 5 of 11
ORDER ON SANCTIONS - 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 First, the volume and severity of Ms. Stewart’s violations show that her submissions were not “subjected to any meaningful scrutiny.” Parker, 2025 WL 4228413, at *8. “Many of the errors were obvious on the face of the document. A competent attorney would have, upon review of the arguments and authority cited, flagged that something was wrong.” Id. But Ms. Stewart admits that she simply did not review the outputs of ChatGPT and Claude because she believed them to be without error. See Dkt. 245 at 7 (“Using automation in this manner gave me confidence that . . . so long as the inputs were reliable, the output would be accurate.”); id. at 15 (“I obtained that characterization from the tool when I asked it to summarize what Dr. Kelman relied upon, and I did not verify it against his report before including it in the motion.”); id. at 17 (“I did not verify the accuracy of the citation information that the tool generated against the actual articles.”). Ms. Stewart’s blind trust in these programs was unfounded. “[A] lawyer using technology in law practice is obliged to do so competently.” Wash. State Bar Ass’n, Advisory Opinion 202505 (Nov. 20, 2025). “[T]he imposition of sanctions for submitting generative-AI hallucinations is so well-documented at this point” that Ms. Stewart’s repeated failure to verify citations generated by these programs “rises to the level of bad faith.” Chaney v. Transdev Servs. Inc., No. 2:24-CV-10761-ODW (AJRX), 2026 WL 1146736, at *6 (C.D. Cal. Apr. 28, 2026). Ms. Stewart even states that she was initially skeptical about using generative artificial intelligence because she had previously heard of an attorney being sanctioned for “us[ing] what was generated without checking it.” Dkt. 245 at 2. And by the time Ms. Stewart filed the offending motions in late 2025, numerous Ninth Circuit courts had issued or threatened sanctions for similar misconduct. See, e.g., Twist It Up, Inc. v. Annie Int’l, Inc., No. SA CV 24-736-DMG (ADSX), 2025 WL 3049841, at *2 (C.D. Cal. June 26, 2025); Ligeri v. Amazon.com Servs. LLC, Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 6 of 11
ORDER ON SANCTIONS - 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 No. 2:25-CV-00764-JHC, 2025 WL 2161497, at *2 (W.D. Wash. July 30, 2025); Oneto v. Watson, 808 F. Supp. 3d 974, 980 (N.D. Cal. Oct. 10, 2025). Ms. Stewart avers that she believed the programs she used would generate correct citations so long as they operated in a “closed system” and referred only to documents she provided. Dkt. 245 at 5–6. But while a closed system might reduce the risk of errors, Ms. Stewart’s belief that it would eliminate that risk was unsupported. See Jonathan Ciottone, et al., A Lawyer’s Guide to Understanding AI Hallucinations in a Closed System, NAT’L L. REV. (June 26, 2024), https://natlawreview.com/article/lawyers-guide-understanding-ai-hallucinations-closed-system (“Improper use of the AI tool, even in a closed system designed for legal use, can lead to illogical outputs or hallucinations.”). In any event, Ms. Stewart’s “failure to make any inquiry at all into the veracity” of the citations “was so ‘completely without merit’ that subjective bad faith may be inferred.” Ramirez v. Humala, No. 24-CV-242 (RPK) (JAM), 2025 WL 1384161, at *2 (E.D.N.Y. May 13, 2025) (quoting In re 60 E. 80th St. Equities, Inc., 218 F.3d 109, 115 (2d Cir. 2000)). Second, the citation errors in the expert evidence are especially flagrant because they indicate Ms. Stewart “may have been so involved with the drafting of expert reports as to ‘jeopardize the experts’ independence.’” Dkt. 209 at 5 (quoting Gerke v. Travelers Cas. Ins. Co. of Am., 289 F.R.D. 316, 326 (D. Or. 2013)); see also McClellan v. I-Flow Corp., 710 F. Supp. 2d 1092, 1118 (D. Or. 2010) (“[E]xpert reports may be discredited if they ‘merely express the opinions of the lawyers who hired them.’” (quoting Trigon Ins. Co. v. United States, 204 F.R.D. 277, 294 (E.D. Va. 2001))). Ms. Stewart now states that she used Claude or ChatGPT to “generate a formatted citation table” for academic articles and “provided the same AI-generated citation table to both experts as an appendix, and neither expert caught the errors in the citation data before signing their reports.” Dkt. 245 at 17–18. Although the Court has already struck these Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 7 of 11
ORDER ON SANCTIONS - 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 expert opinions for other reasons, Dkt. 202 at 17–20, the fact that Ms. Stewart’s faulty citations tainted two expert reports further indicates bad faith. Third, Defendants first raised their concerns about AI hallucinations during oral argument on December 2, 2025, when they provided the Court and Ms. Stewart with a table of allegedly false citations in Plaintiff’s Daubert motion. Dkt. 187 at 48. On December 9, Defendants filed a similar table with respect to Plaintiff’s motion to strike affirmative defenses. Dkt. 191-1. Several citations in these tables were clearly erroneous and ultimately formed the basis for the Court’s order to show cause and the present order. Dkt. 209 at 3–4. Ms. Stewart should have reviewed these citations with urgency and—upon realizing that they were erroneous—revisited all citations that she generated with artificial intelligence, withdrawn or amended all motions containing those citations, and explained her mistakes to the Court. See, e.g., Mavy, 2026 WL 91483, at *7 (finding no bad faith where supervising attorney “immediately pursued an investigation to determine what had happened[,] did not attempt to defend the opening brief but submitted a proposed brief containing correct citations[, and] proactively investigated the contract attorney’s work in different cases and found ‘four other briefs containing erroneous citations’”). Prior to responding to the order to show cause, Ms. Stewart did little to remedy her errors. She submitted an additional hallucinated citation to this Court on December 18, over two weeks after learning that her citations in other briefs might have been erroneous. Dkt. 209 at 5 (citing Dkt. 197). She also filed a “corrected” Daubert motion, which fixed an errant citation but completely changed the meaning of the relevant argument. Id. at 5–6 (citing Dkt. 192 at 2–3). Ms. Stewart now admits that even the corrected motion still contained some of the original false citations that Defendants had highlighted. See Dkt. 245 at 20 (“I also acknowledge that the corrections were incomplete. Several errors identified in the original motion remained Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 8 of 11
ORDER ON SANCTIONS - 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 uncorrected in Dkt. 192, including the Hankey quotation, the ISO accreditation attribution to Dr. Kelman, and the characterization of Dr. Tucker’s opinion in a later section.”). At no point before the Court ordered Ms. Stewart to show cause on February 4, 2026 did she explain the origin of the errant citations. Ms. Stewart’s conduct “exceed[ed] mere ignorance or negligence” and was tantamount to bad faith. Parker, 2025 WL 4228413, at *6. The Court may therefore issue sanctions for Ms. Stewart’s Rule 11 violations on its own motion, and it does so here. B. Sanctions The Court must now determine what types of sanctions against Ms. Stewart are sufficient to deter future, similar misconduct from her and other attorneys. “Rule 11 sanctions are meant to serve several purposes, including (1) deterring future litigation abuse, (2) punishing present litigation abuse, (3) compensating victims of litigation abuse, and (4) streamlining court dockets and facilitating case management.” Wadsworth v. Walmart Inc., 348 F.R.D. 489, 496–97 (D. Wyo. 2025) (quoting White v. Gen. Motors Corp., 908 F.2d 675, 683 (10th Cir. 1990)). Sanctions for fabricated citations may include “(1) monetary sanctions, (2) striking the filing containing erroneous citations, (3) requiring written notification to the client, (4) requiring written notification to judges incorrectly identified as having authored nonexistent cases, and (5) referring attorneys to the appropriate disciplinary body for disciplinary proceedings.” Oneto, 808 F. Supp. 3d at 980 (collecting cases); see, e.g., Chaney, 2026 WL 1146736, at *6 (imposing a $2,500 fine and requiring that the attorney file a declaration explaining the court’s findings of bad faith in his other cases in the district); Tercero v. Sacramento Logistics, LLC, No. 2:24-CV-00953-DC-JDP, 2025 WL 2605020, at *13 (E.D. Cal. Sept. 9, 2025) (imposing a $1,500 fine, requiring that the attorney serve a copy of the court’s sanctions order on her client, and directing the clerk of court to serve the order on the California Bar). Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 9 of 11
ORDER ON SANCTIONS - 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 “[A]ny sanction imposed must be proportionate to the offense and commensurate with principles of restraint and dignity inherent in judicial power.” Oneto, 808 F. Supp. 3d at 980 (quoting Zambrano v. City of Tustin, 885 F.2d 1473, 1480 (9th Cir. 1989)). Courts handling similar cases tend toward leniency where attorneys are forthcoming about their misuse of artificial intelligence and take significant steps to correct their mistakes. See, e.g., Dodge v. FirstService Residential Ariz. LLC, No. CV-24-01550-PHX-SMM, 2025 WL 3653164, at *3 (D. Ariz. Dec. 17, 2025) (imposing no sanctions where an attorney “openly and honestly admitted to his improper use of AI, including the submission of hallucinated citations” and “took the initiative to self-report to the State Bar for his improper use of AI and contend[ed] that he [was] reviewing all other pending matters”). Here, Ms. Stewart’s actions warrant significant sanctions. Ms. Stewart submitted dozens of false citations throughout four briefs and two expert reports. When Defendants identified multiple clearly erroneous citations in December 2025, Ms. Stewart was not proactive in rectifying her mistakes. Her attempt to correct the citations in her Daubert motion was incomplete and confusing, and she did not explain how the errors occurred. After making those corrections, she then submitted a brief containing one additional hallucinated citation. The Court acknowledges that Ms. Stewart responded to the order to show cause by fully admitting fault and explaining how her unverified use of artificial intelligence caused each error. At the same time, “submitting fake citations and filing unverified generative AI outputs . . . prejudices clients, undermines faith in attorneys and the justice system, and compels diversion of public resources from pending cases.” Parker, 2025 WL 4228413, at *9. The Court spent significant time reviewing and addressing Ms. Stewart’s false citations while managing other matters on its docket, including a high volume of time-sensitive immigration habeas matters. Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 10 of 11
ORDER ON SANCTIONS - 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Ms. Stewart’s concessions, while appreciated, cannot replace the time and resources lost. The Court thus finds that the below sanctions are necessary to address Ms. Stewart’s conduct. V. CONCLUSION The Court ORDERS the following sanctions for violation of Federal Rule of Civil Procedure 11(b): 1. Attorney Jocelyn Stewart is personally sanctioned in the amount of $3,000. Within 21 days of the date of this order, Ms. Stewart shall pay this sanction to the Clerk of the Court. The case number and a copy of this order should be included with payment. 2. Ms. Stewart shall include a certification with all subsequent briefing in this case stating that all citations have been verified and accurately reflect the propositions put forth. 3. Ms. Stewart shall provide a copy of this order to Plaintiff Joann LeDoux and file a certification with the Court that she has done so no later than July 28, 2026. 4. The order to show cause issued on February 4, 2026 (Dkt. 209) is hereby discharged. Dated this 24th day of July, 2026. a Tiffany M. Cartwright United States District Judge Case 3:24-cv-05808-TMC Document 265 Filed 07/24/26 Page 11 of 11
ProvenanceKnow exactly where this document came from.Members see the sourcing behind every authority on DocPost — so you can check the record yourself and cite with confidence.Request access