What each tool produced

CASE SUMMARY — Mata v. Avianca, Inc.
Prepared by: Fable 5 (effort: extra high / xhigh)
Source used: "Mata v. Avianca - 678 F.Supp.3d 443 (S.D.N.Y. 2023).pdf" (~/Downloads), 33 PDF pages, reporter pages 443–475.
Source limits: Only the text of the provided PDF was used. Pin cites are to the Federal Supplement, Third Series pagination printed in the PDF. "FOF" = the court's numbered Findings of Fact; "COL" = its numbered Conclusions of Law. The PDF does not appear truncated: the opinion runs from the caption at 443 through Appendix B at 475. (PDF page 1 also contains the tail of an unrelated preceding case, and PDF page 33 the caption of the following case; neither was used.)

======================================================================
1. CASE NAME AND CITATION
======================================================================
Roberto Mata, Plaintiff, v. Avianca, Inc., Defendant. No. 22-cv-1461 (PKC).
Reported at 678 F. Supp. 3d 443 (S.D.N.Y. 2023); running head: "MATA v. AVIANCA, INC. — Cite as 678 F.Supp.3d 443 (S.D.N.Y. 2023)."

======================================================================
2. COURT AND DATE
======================================================================
United States District Court, Southern District of New York (P. Kevin Castel, United States District Judge).
"Opinion and Order on Sanctions," signed June 22, 2023.

======================================================================
3. PROCEDURAL POSTURE
======================================================================
- Nature of the ruling: a trial-court sanctions ruling following an Order to Show Cause and an evidentiary hearing. This is not an appellate decision; the only proceeding "below" is the state-court filing that preceded removal.
- Mata commenced the action on or about February 2, 2022 in the Supreme Court of the State of New York, New York County, alleging a knee injury from a metal serving cart on a flight from El Salvador to JFK. Avianca removed to the S.D.N.Y. on February 22, 2022, asserting federal-question jurisdiction under the Montreal Convention. (FOF 1, at 449.)
- Avianca moved to dismiss on January 13, 2023, arguing the claims were time-barred under the Montreal Convention. (FOF 3, at 449.) On March 1, 2023, Peter LoDuca filed an "Affirmation in Opposition" that cited and quoted the non-existent decisions. (FOF 5, at 450.) Avianca's March 15, 2023 reply stated it could not locate most of the cited cases. (FOF 7, at 450.)
- Orders of April 11 and 12, 2023 directed LoDuca to file an affidavit annexing copies of the cited decisions. (FOF 13–14, at 451–52.) LoDuca's April 25, 2023 Affidavit annexed purported copies or excerpts. (FOF 18, at 452.)
- On May 4, 2023, the Court ordered LoDuca to show cause why he should not be sanctioned under Rule 11(b)(2) and (c), 28 U.S.C. § 1927, and the Court's inherent power. (FOF 42, at 457–58.) A May 26, 2023 supplemental Order directed Steven A. Schwartz and the Levidow Firm to show cause, and directed LoDuca to show cause regarding "the use of a false or fraudulent notarization in the April 25 Affidavit." (FOF 48, at 459.)
- A sanctions hearing was held June 8, 2023; LoDuca and Schwartz testified under oath, and Thomas R. Corvino of the Levidow Firm gave a statement. (FOF 50, at 459.) The opinion sets out Findings of Fact and Conclusions of Law following that hearing. (At 449.)
- Disposition of the underlying motion to dismiss: not stated in the provided text.

======================================================================
4. MATERIAL FACTS
======================================================================
- Schwartz (Levidow Firm) was counsel on the state-court complaint but is not admitted in the S.D.N.Y. and filed no notice of appearance; LoDuca (same firm) appeared after removal, while Schwartz "continued to perform all substantive legal work." (FOF 2, at 449.)
- The March 1 Affirmation in Opposition, signed by LoDuca "under penalty of perjury," was researched and written by Schwartz; LoDuca reviewed it only for style, read none of the cited authorities, and made no inquiry of Schwartz. (FOF 5–6, at 450.)
- Schwartz used ChatGPT, which "fabricated the cited cases"; he had never used it before, believed it was "like a super search engine," had only chatbot-generated excerpts (never full opinions), and his firm's Fastcase account had limited federal access. (FOF 11, 37–40, at 451, 456–57.) He had entered the "Varghese" citation in a free site and "couldn't find it," yet cited it anyway. (FOF 12, at 451.)
- After Avianca's March 15 reply flagged the missing cases, no Respondent sought to withdraw the Affirmation or offered any explanation; LoDuca forwarded the reply to Schwartz without reading it. (FOF 8, 10, at 450–51.)
- In requesting an extension to comply with the April 11–12 Orders, LoDuca told the Court he was "out of the office on vacation," which was false — Schwartz was the one on vacation; the Court found this a knowingly false statement that concealed Schwartz's role. (FOF 16–17, at 452.)
- The April 25 Affidavit was drafted by Schwartz; LoDuca signed it without having seen a draft and without asking a question. It annexed excerpts (not full opinions), stated "Zicherman" could not be located, referred only to an unnamed "online database," and did not disclose ChatGPT. (FOF 18–23, at 452–53.)
- The annexed "Varghese" opinion was not authentic (confirmed by the Eleventh Circuit Clerk); its docket number and reporter citation belong to other cases, its analysis is "gibberish," and it cites further non-existent decisions. Similar defects infected "Miller," "Petersen," "Shaboon," "Martinez," and "Durden." (FOF 24–36, at 453–56.)
- Schwartz's May 25 affidavit was the first acknowledgment that ChatGPT was used; it said ChatGPT was used "to supplement" his research, which he conceded at the hearing was untrue, and his accounts of when he asked ChatGPT whether "Varghese" was real shifted between filings. (FOF 41, 44–47, at 457–59.) Mitigating facts the Court credited: sincere remorse, no financial gain or personal animus, no disciplinary history, significant publicity, the Firm's arranged CLE on technological competence and AI, and Avianca's decision not to seek fees. (COL 30–32, at 466.)

======================================================================
5. ISSUES
======================================================================
(a) Did Respondents violate 18 U.S.C. § 505 (forging a federal judge's signature or a court seal) by citing and submitting fake judicial opinions?
(b) May the Court impose Rule 11 sanctions on Schwartz, who is not admitted in the District and filed no notice of appearance?
(c) Did LoDuca act with subjective bad faith in violating Rule 11(b)(2)?
(d) Did Schwartz act with subjective bad faith in violating Rule 11(b)(2)?
(e) Is the Levidow Firm jointly and severally liable for the Rule 11(b)(2) violations?
(f) Should a separate sanction issue under 28 U.S.C. § 1927?
(g) What sanction is appropriate?

======================================================================
6. HOLDINGS
======================================================================
(a) No — because the fake opinions "do not include any signature or seal," Respondents did not violate § 505, although the Court noted the conduct "raises similar concerns" to those in the forgery case it discussed. (COL 8, at 461.)
(b) Yes — Rule 11(c)(1) authorizes a sanction on "any attorney ... that violated the rule or is responsible for the violation," so the Court may sanction Schwartz. (COL 13, at 462.)
(c) Yes — LoDuca acted in subjective bad faith by signing and filing the Affirmation "after making no 'inquiry,'" swearing to the April 25 Affidavit with no basis, and lying about being on vacation. (COL 23, at 464.)
(d) Yes — Schwartz acted in subjective bad faith because he was aware of a high probability that "Varghese" and "Zicherman" did not exist and "consciously avoided confirming that fact," compounded by the untruthful "supplement" assertion, conflicting accounts, and nondisclosure of ChatGPT. (COL 24, at 464–65.)
(e) Yes — under Rule 11(c)(1) the Firm is jointly and severally liable because it "has not pointed to exceptional circumstances" warranting departure from the rule. (COL 25, at 465.)
(f) No — the Court declined to sanction separately under § 1927 because that statute targets unreasonable and vexatious multiplication of proceedings, and "dilatory tactics and delay were not among" the harms caused. (COL 26, at 465.)
(g) Each Respondent is sanctioned under Rule 11 "and, alternatively, under the inherent power of this Court": notice letters to the client and to each judge falsely named as an author, plus a $5,000 penalty, jointly and severally, paid into the Court Registry. (COL 27, 33–34; Conclusion, at 465–66.)

======================================================================
7. REASONING (≤150 words)
======================================================================
The court applied Rule 11(b)(2), which requires legal contentions warranted by existing law or a nonfrivolous argument after reasonable inquiry, a duty that continues when counsel is "later advocating" the contention (at 461). Acting sua sponte, the court required subjective bad faith, the inherent-power standard; knowing false statements suffice, conscious avoidance equals knowledge, and each lawyer is assessed separately (at 462–64). "A fake opinion is not 'existing law'" (at 461), and Respondents kept advocating the fakes after Avianca's March 15 reply and the April 11–12 Orders. LoDuca made no inquiry before signing, swore to an affidavit he had never seen, and lied about a vacation (at 464). Schwartz knew "Varghese" could not be found, hid his ChatGPT reliance, and falsely called it a "supplement" (at 464–65). The Firm is jointly liable absent exceptional circumstances; § 1927 fails for lack of delay; the sanction is capped at deterrence (at 465–66).

======================================================================
8. DISPOSITION
======================================================================
Sanctions ordered ("SO ORDERED"), pursuant to Rule 11 or, alternatively, the Court's inherent authority (at 466):
(a) within 14 days, Respondents must mail a letter to plaintiff Mata identifying and attaching the Opinion and Order, the June 8, 2023 hearing transcript, and the April 25 Affirmation with exhibits;
(b) within 14 days, Respondents must mail a letter to each judge falsely identified as the author of the fake "Varghese," "Shaboon," "Petersen," "Martinez," "Durden," and "Miller" opinions, with the same attachments plus the fake opinion attributed to that judge;
(c) within 14 days, Respondents must file copies of those letters;
(d) a $5,000 penalty is imposed jointly and severally on Respondents, payable into the Court Registry within 14 days.
Not ordered: attorney's fees (Avianca did not seek them), additional mandatory CLE (redundant of the Firm's own program), or a compelled apology ("a compelled apology is not a sincere apology"). (COL 30–31, 33, at 466.)
Appellate terms (affirmed/reversed/remanded) do not apply; this is an original trial-court order. Disposition of the motion to dismiss: not stated in the provided text.

======================================================================
9. RULE OF LAW
======================================================================
Under Rule 11(b)(2), a non-existent judicial opinion is not "existing law," and citing it supplies no nonfrivolous argument; because the Rule 11 duty continues when a lawyer later advocates the contention, failing to withdraw or correct the filing after being put on notice that the authorities cannot be found is itself sanctionable. Where a court proceeds sua sponte (or under its inherent power), sanctions require subjective bad faith, which may be established by knowingly false statements to the court or by conscious avoidance of a high probability that the cited authorities do not exist; each attorney's knowledge is assessed separately, and, absent exceptional circumstances, the law firm is jointly responsible under Rule 11(c)(1).

======================================================================
10. SEPARATE OPINIONS
======================================================================
None. The provided text contains a single opinion by Judge Castel; no concurrence or dissent appears.

======================================================================
VERIFICATION CHECKLIST
======================================================================
Confirm each of the following against the original opinion before relying on this summary:

Identification and source
[ ] The citation is 678 F. Supp. 3d 443 (S.D.N.Y. 2023), docket No. 22-cv-1461 (PKC), signed June 22, 2023, authored by Judge P. Kevin Castel.
[ ] The "Background," "Holdings," and numbered headnotes at 443–448 are publisher's editorial material (West synopsis and Key Number headnotes), not the court's words. This summary relies on the opinion body (448–466) and appendices; do not cite the headnotes as the court's holding.
[ ] PDF page 1 (left column) is the end of an unrelated preceding case (an FLSA collective-action ruling) and PDF page 33 (bottom) is the caption of the following case (Hermès v. Rothschild); neither is part of Mata.

Procedural facts
[ ] Removal date (Feb. 22, 2022), motion-to-dismiss date (Jan. 13, 2023), Affirmation date (Mar. 1, 2023), Avianca reply (Mar. 15, 2023), Orders of Apr. 11 and 12, 2023, April 25 Affidavit, OSCs of May 4 and May 26, 2023, and hearing of June 8, 2023.
[ ] The opinion does not state how the underlying motion to dismiss was resolved.
[ ] Footnote 4: the April 11 Order's deadline was written as "April 18, 2022," not 2023 (a noted date discrepancy).

Holdings and legal standards
[ ] § 505 holding rests solely on the absence of any signature or seal on the fake opinions (COL 8, at 461).
[ ] Authority to sanction Schwartz rests on the "any attorney ... responsible for the violation" language of Rule 11(c)(1) (COL 13, at 462).
[ ] The court treated the proceeding as sua sponte under Rule 11(c)(3) and therefore applied a subjective-bad-faith standard (COL 11, 14, at 462); it did not apply the objective-unreasonableness standard that applies when an adversary moves under Rule 11(c)(2).
[ ] Conscious-avoidance formulation: "aware of a high probability of the fact in dispute and consciously avoided confirming that fact" (COL 19, at 463, quoting Svoboda).
[ ] Separate-assessment principle: each respondent's knowledge and conduct assessed separately; no imputation (COL 22, at 464, quoting Weddington).
[ ] Sanctions are imposed under Rule 11 "and, alternatively," the inherent power (COL 27, at 465; Conclusion, at 466) — the court did not elect between the two bases.
[ ] § 1927 declined for lack of delay or vexatious multiplication (COL 26, at 465).
[ ] Sanction components and the 14-day deadlines (Conclusion, at 466); $5,000 is a penalty into the Registry, not fees.

Quotations (all under 25 words; confirm exact wording and pin cite)
[ ] "existing rules impose a gatekeeping role on attorneys to ensure the accuracy of their filings" (at 448).
[ ] "A fake opinion is not 'existing law'" (at 461).
[ ] "sua sponte sanctions in those circumstances should issue only upon a finding of subjective bad faith" (at 462, quoting Muhammad).
[ ] "signing and filing that affirmation after making no 'inquiry' was an act of subjective bad faith" (at 464).
[ ] "consciously avoided confirming that fact" (at 465).
[ ] "dilatory tactics and delay were not among them" (at 465).
[ ] "sufficient but not more than necessary to advance the goals of specific and general deterrence" (at 466).
[ ] "a compelled apology is not a sincere apology" (at 466).

Ambiguities flagged (not resolved in this summary)
[ ] Footnote 3 (at 451): the court says Schwartz's testimony that he "couldn't find" Varghese "appears to acknowledge" he knew this before the March 1 Affirmation, but "also could refer to the April 25 Affidavit"; the court resolves it only as "[e]ither way, he knew before making a submission." The summary does not fix the timing.
[ ] Footnote 5 (at 452): Schwartz's June 6 declaration says the April 25 Affidavit was signed in his (Schwartz's) office; his hearing testimony said LoDuca's office. The court notes the conflict without resolving it.
[ ] Notarization ground: the May 26 Order directed LoDuca to show cause regarding "a false or fraudulent notarization" in the April 25 Affidavit (FOF 48, at 459). The Conclusions of Law contain no express finding on that ground; the only related statement is that the Firm intends to hold notarization training (COL 30, at 466). Whether the ground was abandoned, subsumed, or resolved is not stated in the provided text.
[ ] Footnote 1 (at 448): the court notes that counsel for Schwartz and the Firm initially included three of the fake cases in a Table of Authorities in their OSC response brief, which counsel "promptly caught and corrected"; the court characterizes this as an "innocent mistake." Confirm this was not treated as sanctionable conduct.
[ ] The phrase "Respondents' reliance on fakes cases" (COL 26, at 465) appears as printed; the summary reads it as "fake cases."

Scope limits of this summary
[ ] Facts in Section 4 are limited to those the court relied on for bad faith, firm liability, and the sanction; the detailed catalog of defects in each fake opinion (FOF 24–35) is summarized, not reproduced.
[ ] The Rule of Law (Section 9) is phrased no more broadly than the court's own statements at 461–65; it should not be read as a holding on the permissible use of AI tools generally — the court expressly said "there is nothing inherently improper about using a reliable artificial intelligence tool for assistance" (at 448).

Mata v. Avianca, Inc. — Judicial Opinion Summary
Source limited to: Mata v. Avianca - 678 F.Supp.3d 443 (S.D.N.Y. 2023).pdf
Prepared as a senior-associate working summary. Not a substitute for the opinion.

1. Case name and citation
Roberto MATA, Plaintiff, v. AVIANCA, INC., Defendant.
22-cv-1461 (PKC)
Cite as 678 F.Supp.3d 443 (S.D.N.Y. 2023)
Running caption as printed: MATA v. AVIANCA, INC.

2. Court and date
United States District Court, S.D. New York.
CASTEL, United States District Judge (the West synopsis also names “P. Kevin Castel, J.”).
Signed June 22, 2023.
Findings of Fact and Conclusions of Law follow a hearing of June 8, 2023.

3. Procedural posture
This is a district-court Opinion and Order on Sanctions, not an appellate decision. Roberto Mata filed a Verified Complaint in the Supreme Court of the State of New York, New York County. Avianca, Inc. removed the action to this Court on February 22, 2022. Avianca later moved to dismiss. Plaintiff’s counsel filed an Affirmation in Opposition citing purported authorities. After Avianca’s reply questioned those authorities, the Court ordered production of the cited decisions, received an April 25, 2023 affidavit annexing purported copies or excerpts, issued Orders to Show Cause (May 4, 2023, as to Peter LoDuca; May 26, 2023, supplemental as to Steven A. Schwartz, the Levidow Firm, and additional grounds including notarization), and held a June 8, 2023 sanctions hearing. This opinion decides those show-cause proceedings. The provided text does not decide Avianca’s motion to dismiss.

4. Material facts
- Mata alleged injury when a metal serving cart struck his left knee on a flight from El Salvador to John F. Kennedy Airport. Avianca removed, asserting federal-question jurisdiction under the Montreal Convention. (678 F. Supp. 3d at 448–49.)
- Steven A. Schwartz of Levidow, Levidow & Oberman P.C. had been listed on the state-court complaint but is not admitted in this District. Peter LoDuca of that firm filed the federal notice of appearance; Schwartz continued to perform all substantive legal work. (Id. at 449.)
- After Avianca moved to dismiss Mata’s claims as time-barred under the Montreal Convention, Schwartz used ChatGPT to generate the legal authorities supporting an argument that the federal bankruptcy stay tolled that limitations period. The firm’s Fastcase account had limited federal access; Schwartz had not previously used ChatGPT. LoDuca signed and filed a March 1, 2023 “Affirmation in Opposition” citing purported Federal Reporter, Federal Supplement, and Westlaw decisions. LoDuca did not review any cited judicial authorities and made no inquiry of Schwartz about the research. (Id. at 449–50, 456–57.)
- Avianca’s March 15, 2023 reply stated that counsel had been unable to locate most of the cited cases and that the few located did not stand for the cited propositions. No respondent withdrew the March 1 Affirmation or explained how a purported Federal Reporter or Federal Supplement decision could not be found. (Id. at 450.)
- The Court ordered LoDuca to annex copies of identified cited decisions (Orders of April 11 and 12, 2023). LoDuca obtained an extension by a letter stating that “the undersigned is currently out of the office on vacation,” a statement the Court found he knew was false; Schwartz, not LoDuca, was away. Schwartz drafted the April 25, 2023 affidavit; LoDuca signed it under oath without authoring it, without a prior draft, and without questioning Schwartz. The affidavit annexed excerpts of now-admitted fake opinions, did not name the “online database,” and stated that “Zicherman v. Korean Air Lines Co., Ltd., 516 F.3d 1237 (11th Cir. 2008)” could not be located. (Id. at 451–52.)
- Schwartz testified that he entered a citation to “Varghese” on a free internet site and “couldn’t find it,” yet the case was cited to the Court. Respondents first acknowledged that ChatGPT had generated the bogus cases in Schwartz’s May 25, 2023 affidavit, after the May 4 Order to Show Cause. The Court found Schwartz’s written assertion that ChatGPT merely “supplement[ed]” other research misleading; at the hearing he agreed ChatGPT “became [his] last resort” and “was [his] research.” At no time did any respondent write to withdraw the March 1 Affirmation or advise the Court it could no longer be relied upon. (Id. at 451, 457–59.)
- At the June 8, 2023 hearing, LoDuca and Schwartz expressed remorse. Thomas R. Corvino, described as the sole equity partner, described firm remedial steps, including expanded Fastcase access and planned mandatory CLE on technological competence, AI tools, and notarization practices. Avianca did not seek reimbursement of attorneys’ fees or expenses. The Court found the fake cases were not submitted for any respondent’s financial gain and were not done out of personal animus; respondents did not have a history of disciplinary violations. (Id. at 459, 466.)

5. Issue(s)
A. Did the citation and submission of the fake opinions violate 18 U.S.C. § 505 (forgery of a United States judge’s signature or a federal court’s seal)?
B. May the Court impose a Rule 11 sanction on Schwartz even though he is not admitted to practice in this District and did not file a notice of appearance?
C. Because the Court proceeded sua sponte, did LoDuca act with subjective bad faith in violating Rule 11?
D. Because the Court proceeded sua sponte, did Schwartz act with subjective bad faith in violating Rule 11?
E. Absent exceptional circumstances, is the Levidow Firm jointly and severally liable under Rule 11(c)(1) for the Rule 11(b)(2) violations of LoDuca and Schwartz?
F. Should the Court separately impose a sanction under 28 U.S.C. § 1927?
G. What sanction is “limited to what suffices to deter repetition of the conduct or comparable conduct by others similarly situated”?

6. Holding
A. No. The fake opinions “do not include any signature or seal, and the Court therefore concludes that Respondents did not violate section 505.” (Id. at 461.)
B. Yes. Rule 11(c)(1) permits a sanction on “any attorney … that violated the rule or is responsible for the violation,” so the Court had authority to sanction Schwartz. (Id. at 462.)
C. Yes. LoDuca acted with subjective bad faith by filing the March 1 Affirmation after making no inquiry into the cited law, by swearing to the April 25 Affidavit with no basis for doing so, and by lying that he was going on vacation to obtain an extension. (Id. at 464–65.)
D. Yes. Schwartz acted with subjective bad faith by remaining silent after being unable to find “Varghese” (and offering no explanation for inability to find “Zicherman”), by consciously avoiding confirmation of a high probability that those decisions did not exist, by untruthfully describing ChatGPT as a mere “supplement,” by giving conflicting accounts of his ChatGPT queries about whether “Varghese” was real, and by failing to disclose reliance on ChatGPT in the April 25 Affidavit. (Id. at 464–65.)
E. Yes. The firm “has not pointed to exceptional circumstances that warrant a departure from Rule 11(c)(1).” (Id. at 465.)
F. No. Respondents’ reliance on fake cases “has caused several harms but dilatory tactics and delay were not among them.” (Id. at 465.)
G. A $5,000 penalty paid into the Registry of the Court, jointly and severally, plus required notice letters to Mata and to judges whose names were used, is “sufficient but not more than necessary to advance the goals of specific and general deterrence.” Further mandatory education would be redundant; an award of opposing counsel’s fees is “not warranted”; a compelled apology will not be required. Sanctions are imposed under Rule 11 and, alternatively, the Court’s inherent power. (Id. at 465–66.)

7. Reasoning
Rule 11(b)(2) certifies, after reasonable inquiry, that legal contentions are warranted by existing law or a nonfrivolous argument to change it. A fake opinion is not existing law; citing one abuses the adversary system. Sua sponte Rule 11 sanctions require subjective bad faith, including conscious avoidance, assessed separately as to each lawyer. LoDuca signed the March 1 opposition and April 25 affidavit without reading cited authorities, knowing Schwartz lacked federal experience, and lied that he was on vacation to get more time. Schwartz treated ChatGPT as his research source, could not find Varghese, withheld that fact, kept advocating the citations after Avianca and the Court could not locate them, and falsely described ChatGPT as a supplement. The firm showed no exceptional circumstances under Rule 11(c)(1). Section 505 did not apply (no signature or seal). Section 1927 did not apply (no delay). A $5,000 penalty was the least severe deterrent.

8. Disposition
Not an appellate affirmance, reversal, remand, or rendition. The West synopsis states: “Ordered accordingly.”
The Court ordered these sanctions pursuant to Rule 11 or, alternatively, its inherent authority (id. at 466):
(a) Within 14 days, Respondents shall mail Mata a letter identifying and attaching this Opinion and Order, a transcript of the June 8, 2023 hearing, and “a copy of the April 25 Affirmation, including its exhibits.”
(b) Within 14 days, Respondents shall mail a letter to each judge falsely identified as author of the fake “Varghese,” “Shaboon,” “Petersen,” “Martinez,” “Durden,” and “Miller” opinions, identifying and attaching this Opinion and Order, the June 8 transcript, and a copy of the April 25 Affirmation including the fake “opinion” attributed to the recipient judge.
(c) Within 14 days, Respondents shall file copies of those letters.
(d) A penalty of $5,000 is jointly and severally imposed on Respondents, payable into the Registry of the Court within 14 days.
SO ORDERED.

9. Rule of law
A fake opinion is not “existing law,” and citation to a fake opinion does not provide a non-frivolous argument for extending, modifying, or reversing existing law or for establishing new law; attempting to persuade a court or oppose an adversary by relying on fake opinions is sanctionable abuse of the adversary system. When a court proceeds sua sponte, Rule 11 sanctions require subjective bad faith as to each lawyer, which may be shown by conscious avoidance; absent exceptional circumstances, the lawyer’s firm is jointly responsible under Rule 11(c)(1). The opinion also states that “there is nothing inherently improper about using a reliable artificial intelligence tool for assistance,” but “existing rules impose a gatekeeping role on attorneys to ensure the accuracy of their filings.” (Id. at 448, 461, 462, 464–65.)

10. Separate opinions
None.

Verification checklist
Confirm each of the following against the original opinion before relying on this summary.

Caption, citation, court, date
- Case name appears as Roberto MATA v. AVIANCA, INC., docket 22-cv-1461 (PKC), “Cite as 678 F.Supp.3d 443 (S.D.N.Y. 2023).”
- Deciding court is the United States District Court, S.D. New York; judge is Castel (West synopsis: P. Kevin Castel, J.).
- The opinion is “Signed June 22, 2023,” after a June 8, 2023 hearing.
- This document is an “OPINION AND ORDER ON SANCTIONS,” not an appellate disposition.

What this opinion does not decide
- The provided text does not state the Court’s ruling on Avianca’s motion to dismiss.
- The provided text does not state any disciplinary-referral order, even though advisory-committee notes quoted in the opinion list referral as a possible type of Rule 11 sanction.
- The May 26 Order to Show Cause directed a response regarding “use of a false and fraudulent notarization in the April 25 Affidavit.” The Conclusions of Law do not state a finding that the notarization was false or fraudulent. Do not treat that OSC allegation as a holding.
- The opening pages of the supplied PDF contain leftover text from a different case (an FLSA collective-action order) and the closing pages begin Hermès International v. Rothschild, 678 F. Supp. 3d 475. Those materials are not part of this opinion.

Key facts to re-check
- Removal date (February 22, 2022); Montreal Convention as the asserted jurisdictional basis; alleged injury (metal serving cart; left knee; El Salvador to JFK).
- Identity of respondents: Peter LoDuca; Steven A. Schwartz; Levidow, Levidow & Oberman P.C.
- Schwartz not admitted in the District; LoDuca as attorney of record; Schwartz as author of the substantive work, including the March 1 Affirmation and the April 25 affidavit.
- ChatGPT as the source of the fabricated authorities; Fastcase limitations as Schwartz’s stated reason for turning to ChatGPT.
- Avianca reply dated March 15, 2023, asserting inability to locate most cited cases.
- Court Orders of April 11 and 12, 2023, requiring production of named decisions, with Rule 41(b) dismissal threatened for noncompliance with the April 11 Order.
- LoDuca’s extension letter claiming he was “out of the office on vacation,” found knowingly false.
- First acknowledgment of ChatGPT on May 25, 2023, after the May 4 OSC; supplemental OSC May 26, 2023.
- No written withdrawal of the March 1 Affirmation.
- Avianca did not seek fees; fake cases not submitted for financial gain or personal animus; no history of disciplinary violations, as the Court found.

Ambiguities and internal inconsistencies in the opinion (do not pick a reading)
- Timing of Schwartz’s inability to find “Varghese”: footnote 3 states that his testimony “appears to acknowledge” knowledge before the March 1 Affirmation, but “[h]is answer also could refer to the April 25 Affidavit.” The Court then said that either way he knew before a submission that the full text could not be found and kept silent. Report both the ambiguity and the Court’s “either way” sentence; do not choose one timing.
- Conflicting accounts of where the April 25 affidavit was signed (LoDuca’s office versus Schwartz’s office) appear in footnote 5.
- The Court found not credible Schwartz’s claim that he thought “F.3d” meant “federal district, third department,” and noted later contrary testimony (id. at 453 n.6).
- Spelling of one fake case varies: the April 11 Order lists “Peterson v. Iran Air”; later findings use “Petersen.”
- The decretal paragraph refers to “the April 25 Affirmation”; the Findings of Fact call that filing the “April 25 Affidavit.”
- Footnote 4 notes the April 11 Order directed filing by “April 18, 2022, not 2023.”

Holdings and scope (do not broaden)
- Section 505: no violation because no signature or seal; the Court only “notes” similar integrity concerns.
- Rule 11(b)(2): a fake opinion is not existing law; later advocacy / failure to correct remains subject to Rule 11.
- Mens rea: because sanctions were sua sponte, the Court required subjective bad faith (not the objective unreasonableness standard that would apply after a safe-harbor motion). Knowledge was assessed separately as to each lawyer; “principles of imputation of knowledge do not apply.”
- Conscious avoidance, not mere negligence, was the Court’s route to Schwartz’s (and, as to non-inquiry, LoDuca’s) bad faith.
- Firm liability rests on Rule 11(c)(1) “absent exceptional circumstances,” which the firm did not show.
- Section 1927 was declined solely on the Court’s finding that the harms did not include dilatory tactics and delay.
- Inherent authority is stated as an alternative basis for the same sanctions, not as a separately itemized additional penalty.
- The Court did not hold that use of AI is inherently improper; it distinguished a “reliable” AI tool from these respondents’ failure to gatekeep accuracy.
- The $5,000 figure is a penalty paid into the Registry of the Court, jointly and severally, for deterrence; the Court also required notice letters and declined a compelled apology and a fee-shifting award.

Quotations used in this summary (confirm wording and page)
- “did not violate section 505” (461)
- “has not pointed to exceptional circumstances that warrant a departure from Rule 11(c)(1)” (465)
- “has caused several harms but dilatory tactics and delay were not among them” (465)
- “sufficient but not more than necessary to advance the goals of specific and general deterrence” (466)
- “there is nothing inherently improper about using a reliable artificial intelligence tool for assistance” (448)
- “existing rules impose a gatekeeping role on attorneys to ensure the accuracy of their filings” (448)
- “A fake opinion is not ‘existing law’” (461)
- “the undersigned is currently out of the office on vacation” (451)

Names of fake decisions the Court treated as ChatGPT-generated and nonexistent
- Respondents acknowledged as generated by ChatGPT and nonexistent: “Varghese,” “Miller,” “Petersen,” “Shaboon,” “Martinez,” and “Durden.” (Id. at 456.)
- “Zicherman” as cited (516 F.3d 1237 (11th Cir. 2008)) was ordered produced and was reported as unable to be located; the Court described that Federal Reporter citation as not existing as cited. Do not equate that finding with the six-name acknowledgment sentence without checking the surrounding paragraphs.
- The April 11 Order also listed Ehrlich v. American Airlines, Inc., 360 N.J. Super. 360 (App. Div. 2003), and In re Air Crash Disaster Near New Orleans, LA, 821 F.2d 1147, 1165 (5th Cir. 1987). The provided text does not, in the later acknowledgment sentence, list those two among the ChatGPT-generated fakes. Do not assume they were fake.

Sanctions decretal terms
- 14-day deadlines; first-class mail; addressees (Mata; each judge named as author of the six listed fakes); attachments required; $5,000 to the Registry; joint and several liability of LoDuca, Schwartz, and the Levidow Firm.
- Confirm whether the Court required letters about all fake opinions or only the six named in decretal paragraph (b).

1. Case name and citation — MATA v. AVIANCA, INC., 678 F.Supp.3d 443 (S.D.N.Y. 2023).

2. Court and date — United States District Court, S.D. New York; signed June 22, 2023.

3. Procedural posture — Roberto Mata filed a personal-injury action against Avianca in New York state court. Avianca removed the action to federal court under the Montreal Convention and moved to dismiss the claims as time-barred. Mata's attorneys filed an opposition relying on nonexistent judicial opinions generated by ChatGPT. After Avianca questioned the authorities and the court ordered counsel to produce them, counsel submitted purported copies or excerpts of fake opinions. The court issued show-cause orders, received written responses, and held a sanctions hearing on June 8, 2023. The disposition of Avianca's motion to dismiss is not stated in the provided text.

4. Material facts —
• Steven A. Schwartz, who was not admitted in the district, performed the substantive legal work; Peter LoDuca appeared as counsel of record, signed, and filed the submissions.
• Schwartz used ChatGPT as the sole source of his substantive opposition after obtaining no useful results through Fastcase; ChatGPT fabricated cases, quotations, and citations.
• LoDuca filed the March 1, 2023 Affirmation in Opposition under penalty of perjury without reading any cited case or otherwise investigating whether its legal assertions were supported.
• Avianca's March 15 reply identified seven authorities it could not locate, but no respondent withdrew the affirmation or explained the citations.
• After the court ordered production of the cited cases, Schwartz prepared and LoDuca swore to the April 25 Affidavit, which attached excerpts of fake opinions, omitted meaningful disclosure of their source, and admitted that one cited decision could not be located.
• LoDuca knowingly stated that he was on vacation to secure an extension when Schwartz, the undisclosed author, was actually away; LoDuca also consciously avoided reading Avianca's warning and independently checking the authorities.
• Schwartz knew that identified cited cases could not be found, consciously avoided confirming their nonexistence, misleadingly described ChatGPT as supplementing other research, and gave conflicting accounts about when he questioned whether a case was real.
• The fake opinions contained no forged judicial signature or court seal; the firm later arranged training on technological competence, artificial intelligence, and notarization, and Avianca did not seek fees or expenses.

5. Issue(s) —
1. Did submitting the fake opinions violate 18 U.S.C. § 505?
2. Could the court sanction Schwartz under Rule 11 although he was not admitted in the district and had not appeared?
3. Did LoDuca act with the subjective bad faith required for sua sponte sanctions?
4. Did Schwartz act with the subjective bad faith required for sua sponte sanctions?
5. Was the Levidow Firm jointly responsible for its attorneys' Rule 11 violations?
6. Did the conduct warrant a separate sanction under 28 U.S.C. § 1927?
7. What sanctions were sufficient to deter repetition of the conduct or comparable conduct?

6. Holding —
1. No; section 505 was not violated because the fake opinions contained no forged judicial signature or court seal.
2. Yes; Rule 11 authorized sanctions against any attorney who violated the rule or was responsible for the violation.
3. Yes; LoDuca acted with subjective bad faith by making no inquiry, swearing without a factual basis, consciously avoiding warnings, and knowingly making a false statement to obtain an extension.
4. Yes; Schwartz acted with subjective bad faith through conscious avoidance and false or misleading statements concerning the authorities and his research.
5. Yes; the Levidow Firm was jointly and severally responsible because it identified no exceptional circumstances under Rule 11(c)(1).
6. No; although the fake cases caused harms, the conduct did not involve the dilatory tactics or delay addressed by section 1927.
7. A $5,000 joint-and-several penalty and notice to Mata and the judges falsely identified as authors were sufficient but not more than necessary for deterrence.

7. Reasoning — Rule 11 requires reasonable inquiry that filed or later-advocated legal contentions are supported by existing law or a nonfrivolous argument; sua sponte sanctions require subjective bad faith, provable by conscious avoidance. Fake opinions satisfy neither permissible basis. LoDuca made no inquiry, swore without a basis, ignored warnings, and lied about his vacation to obtain an extension. Schwartz knew citations could not be found, avoided confirmation, concealed ChatGPT's role, and gave contradictory accounts. Continuing to advocate after Avianca's warning and the court's orders established bad faith, not negligence. Rule 11 reached Schwartz as an attorney responsible for the violation and made the firm jointly responsible absent exceptional circumstances. Section 505 did not apply because the opinions lacked forged signatures or seals; section 1927 did not apply because the misconduct did not cause delay. Deterrence justified notice and a $5,000 joint penalty.

8. Disposition — The court sanctioned LoDuca, Schwartz, and the Levidow Firm under Rule 11 and, alternatively, its inherent authority. It ordered them, within 14 days, to send specified materials to Mata and each judge falsely identified as an author, file copies of those letters, and pay a $5,000 joint-and-several penalty into the court registry. The court declined a separate section 1927 sanction. The disposition of the underlying motion to dismiss is not stated in the provided text.

9. Rule of law — In a sua sponte Rule 11 proceeding, subjective bad faith may be established by conscious avoidance; reliance on fake opinions, followed by continued advocacy after warnings, is sanctionable because fake opinions are neither existing law nor a nonfrivolous basis for changing the law. Rule 11 reaches an attorney responsible for a violation even if the attorney is not admitted or has not appeared, and the attorney's firm is jointly responsible absent exceptional circumstances.

10. Separate opinions — none.

Verification checklist
• Confirm the exact case caption, reporter citation, deciding court, and signature date.
• Confirm that Mata filed in New York state court, Avianca removed under the Montreal Convention, and Avianca moved to dismiss on limitations grounds.
• Confirm that the provided text does not state the disposition of the underlying motion to dismiss.
• Confirm Schwartz's and LoDuca's respective roles, including Schwartz's lack of admission in the district and LoDuca's status as counsel of record.
• Confirm that Schwartz used ChatGPT as the source of the substantive legal arguments after Fastcase produced no useful research and that ChatGPT fabricated the cited authorities.
• Confirm the contents and sworn nature of the March 1 Affirmation in Opposition and LoDuca's lack of inquiry before signing and filing it.
• Confirm Avianca's March 15 warning, the court's April production orders, and respondents' failure to withdraw or correct the March 1 submission.
• Confirm the contents, preparation, execution, and defects of the April 25 Affidavit, including the submission of excerpts and the inability to locate the purported Zicherman authority.
• Confirm the factual bases for the finding of LoDuca's subjective bad faith, including conscious avoidance, lack of inquiry, unsupported sworn statements, and the false vacation representation.
• Confirm the factual bases for the finding of Schwartz's subjective bad faith, including unlocatable cases, conscious avoidance, nondisclosure of ChatGPT, the claim that ChatGPT supplemented other research, and conflicting accounts.
• Confirm that no fake opinion contained a forged judicial signature or court seal and that the court therefore found no section 505 violation.
• Confirm the court's Rule 11 standards governing reasonable inquiry, frivolous legal contentions, later advocacy, sua sponte sanctions, subjective bad faith, and conscious avoidance.
• Confirm the court's authority under Rule 11(c)(1) to sanction an attorney responsible for a violation despite lack of admission or appearance.
• Confirm the separate assessment of each attorney's knowledge and conduct and the prohibition against imputing one attorney's knowledge to another.
• Confirm the Levidow Firm's joint-and-several responsibility and the absence of exceptional circumstances under Rule 11(c)(1).
• Confirm that the court alternatively relied on its inherent authority and that this authority also required bad faith.
• Confirm that the court declined a separate section 1927 sanction because dilatory tactics and delay were not among the harms caused.
• Confirm the deterrence factors, including the firm's remedial training, Avianca's decision not to seek fees, respondents' remorse, lack of financial gain or personal animus, disciplinary history, and assessed likelihood of recurrence.
• Confirm the exact sanctions, recipients, required attachments, 14-day deadlines, filing requirement, $5,000 amount, joint-and-several liability, and payment into the court registry.
• Confirm that the opinion contains no concurrence or dissent.

CoCounsel echoes the prompt back before its answer. The summary begins at “Case name and citation.”

COCOUNSEL

Result generated by AI Assistant
Result created: August 21, 2026
Prompt: Summarize the judicial opinion Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023) using ONLY the text of the opinion, in exactly this structure:
Case name and citation — as they appear in the opinion.
Court and date — the deciding court and the date of decision.
Procedural posture — how the case arrived at this court, and what happened below.
Material facts — 5-8 bullets, limited to facts the court treated as relevant to its holding.
Issue(s) — each stated as a single question.
Holding — the court’s answer to each issue, one sentence each.
Reasoning — the court’s actual chain of logic, 150 words maximum. Identify the rule the court applied and how it applied that rule to these facts.
Disposition — affirmed, reversed, remanded, rendered, etc.
Rule of law — state the rule this case stands for in one or two sentences, phrased so it could be dropped into a brief.
Separate opinions — summarize any concurrence or dissent in one sentence each, or state “none.”
Constraints: - Quote directly only where the court’s exact language matters. Keep quotations under 25 words, in quotation marks, with a pin cite if the opinion supplies page numbers. - Do not characterize the holding more broadly than the court itself did. - Do not resolve ambiguity in the opinion by picking the more likely reading; flag the ambiguity instead. - Some citations discussed in this opinion are fabricated and are the subject of the court’s analysis. Do not present any such citation as valid authority. - After the summary, add a section titled “Verification checklist” listing every proposition a reader must confirm against the original opinion before relying on this summary.
Resources: Mata v. Avianca, Inc.

Result:
Case name and citation: Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023).
Court and date: United States District Court, S.D. New York, June 22, 2023.
Procedural posture: The case involved a motion for sanctions against attorneys who submitted non-existent judicial opinions in opposition to a motion to dismiss.
Material facts:
Roberto Mata filed a complaint alleging injury during a flight. 
Avianca moved to dismiss, claiming the action was time-barred under the Montreal Convention. 
Attorneys submitted an affirmation citing fake cases generated by ChatGPT. 
The court ordered the attorneys to produce the cited cases, which they failed to do. 
The attorneys did not withdraw the affirmation or correct the record after being informed of the non-existence of the cases. 
The attorneys falsely claimed to have verified the cases through reliable sources. 
Issue(s): Did the attorneys violate Rule 11 by submitting and failing to correct an affirmation citing non-existent cases?
Holding: Yes, the attorneys violated Rule 11 by submitting and failing to correct the affirmation, acting in subjective bad faith. 
Reasoning: The court applied Rule 11, which requires attorneys to ensure that legal contentions are warranted by existing law. The attorneys failed to conduct a reasonable inquiry into the validity of the cases cited, relied solely on ChatGPT without verification, and did not correct the record after being alerted to the issue. Their actions constituted subjective bad faith, warranting sanctions. 
Disposition: Sanctions were imposed on the attorneys and their law firm. 
Rule of law: Under Rule 11, attorneys must conduct a reasonable inquiry into the validity of legal contentions and correct any false statements. Failure to do so, especially when relying on unverified sources, can result in sanctions for subjective bad faith.
Separate opinions: None.

Verification checklist: - Confirm the procedural posture regarding the motion for sanctions. - Verify the material facts, especially the use of ChatGPT and the failure to correct the record. - Check the court’s application of Rule 11 and its reasoning for finding subjective bad faith. - Ensure the disposition accurately reflects the court’s decision to impose sanctions. - Confirm there are no separate opinions.
Generated by AI and may contain mistakes. Not legal advice. A qualified professional must verify accuracy and legal compliance.

MEMORANDUM
To: Supervising Partner
From: Senior Associate
Date: August 21, 2026
Re: Meridian Freight Systems, Inc. - Dana Whitfield Noncompetition Covenant
I. Question Presented
Whether Texas law enforces Whitfield's 24-month, nationwide, any-competitor covenant, signed at the start of at-will employment requiring pricing and roadmap access, or instead reforms it.
II. Brief Answer
Probably not as written, but probably in narrowed form. Meridian's performed promise of confidential-information access likely makes the covenant ancillary despite at-will employment; the nationwide, any-capacity restraint exceeds Meridian's three-state business and Whitfield's work. If ancillary, the covenant must be reformed rather than voided.
III. Facts
Meridian, a Houston logistics-software company, hired Whitfield as an at-will Senior Solutions Engineer on March 3, 2023. Before working, she signed an Employment Agreement in which Meridian promised access to confidential and proprietary customer lists, pricing models, and product roadmaps, without specifying timing. Her duties required configuring pricing models for named accounts and attending roadmap sessions. The agreement bars her, for 24 months after separation, from engaging in, owning, managing, or providing services to any competing business anywhere in the United States. Whitfield resigned June 12, 2026 and accepted a Solutions Architect position with an Austin competitor. Meridian's customers are only in Texas, Louisiana, and Oklahoma.
Facts still needed
Whitfield's accounts and territory; information received and its current sensitivity; any nondisclosure or return terms; and her new duties, customers, and territory.
IV. Discussion
A court likely would find the covenant ancillary but overbroad, then reform it.
The assignment's statutory premise requires correction. Section 15.05 generally prohibits restraints of trade; Sections 15.50-.52 are the Texas Covenants Not to Compete Act. Tex. Bus. & Com. Code §§ 15.05, 15.50-.52 [VERIFY]. Binding Section 15.50(a) requires an otherwise enforceable agreement and reasonable limits on time, geography, and activity no broader than necessary to protect goodwill or another business interest; Meridian bears proof in this personal-services agreement. Id. §§ 15.50(a), 15.51(b) [VERIFY].
Binding Texas Supreme Court authority holds that an at-will employer's promise to provide confidential information supports a covenant once performed, although initially illusory. Alex Sheshunoff Mgmt. Servs., L.P. v. Johnson, 209 S.W.3d 644, 648-51, 655 (Tex. 2006) [VERIFY]. Work requiring confidential information can imply that promise; delivery completes the agreement. Mann Frankfort Stein & Lipp Advisors, Inc. v. Fielding, 289 S.W.3d 844, 849-52 (Tex. 2009) [VERIFY]. Reasonableness remains central. Marsh USA Inc. v. Cook, 354 S.W.3d 764, 777-78 (Tex. 2011) [VERIFY]. Persuasively, Houston's First Court treats the employee's work territory as the usual reasonable area and affirms narrowing overbroad geography. Butler v. Arrow Mirror & Glass, Inc., 51 S.W.3d 787, 793-94 (Tex. App.-Houston [1st Dist.] 2001, no pet.) [VERIFY].
Meridian expressly promised access, and Whitfield's pricing and roadmap duties indicate performance, creating a protectable confidentiality interest. Signing before work and at-will status are weak defenses. But a United States ban bears no apparent relationship to a company serving only three states. Prohibiting ownership or every service for any competitor also reaches passive investment and work unrelated to Whitfield's role. The two-year term may be supportable only if Meridian proves the information remains sensitive that long.
Meridian's strongest response is that logistics software and roadmaps have national competitive utility and that Solutions Engineer and Solutions Architect roles overlap. That supports some restraint, particularly in Austin, but not a nationwide, any-capacity ban without evidence of broader markets and duties.
Because the confidentiality exchange likely satisfies the threshold, Section 15.51(c) directs the court to reform geography, activities, and, if necessary, duration, and enforce only the narrowed covenant. Tex. Bus. & Com. Code § 15.51(c) [VERIFY]. Pre-reformation damages are unavailable; relief is injunctive, and overreaching can expose Meridian to Whitfield's fees under the statute. Id. [VERIFY].
V. Conclusion
Meridian should not demand nationwide compliance. It should preserve access records, identify Whitfield's accounts and sensitive information, compare both roles, and propose a restriction limited to her territory or relevant customers and substantially similar duties. If litigation is necessary, request narrow injunctive relief and reformation; risks are proof of confidential disclosure, overbreadth, no pre-reformation damages, and statutory fees. This preliminary memorandum is not verified legal advice.

MEMORANDUM

    To: Supervising PartnerFrom: Senior AssociateDate: August 21, 2026Re: Meridian Freight Systems, Inc. - Dana Whitfield Noncompetition Covenant

    
    I. Question Presented

    Under Texas law, may Meridian enforce or obtain reformation of Whitfield's two-year nationwide ban on serving any competitor when Meridian promised and her duties required access to confidential pricing and roadmap information, but Meridian serves customers only in Texas, Louisiana, and Oklahoma?

    
    
    II. Brief Answer

    Probably not as written, but likely after reformation if Meridian proves it provided the promised, job-required confidential information. That performance can support an ancillary agreement despite Whitfield's at-will status; however, banning every service to every competitor nationwide exceeds Meridian's demonstrated three-state interest and Whitfield's activities. If the covenant is ancillary, the court must reform unreasonable limits rather than void it.

    
    
    III. Facts

    Meridian, a Houston logistics-software company, hired Whitfield as an at-will Senior Solutions Engineer on March 3, 2023. Before beginning work that day, she signed an Employment Agreement. Meridian promised access to confidential customer lists, pricing models, and product roadmaps, without specifying when. Her duties required configuring pricing models for named accounts and attending internal roadmap sessions. The covenant bars her, for 24 months after separation, from engaging in, owning, managing, or providing services to any competing business anywhere in the United States. Meridian's customers are only in Texas, Louisiana, and Oklahoma. Whitfield resigned June 12, 2026 and accepted a Solutions Architect role with an Austin competitor.

    
    Facts still needed

    Whether Whitfield actually received each promised category; the information's confidentiality, age, and continuing competitive value; her accounts and territory; her new duties; and whether the Agreement contains nondisclosure, severability, or reformation provisions.

    
    
    
    IV. Discussion

    The assignment's statutory premise needs correction: section 15.05 generally prohibits restraints of trade; sections 15.50-.52 specifically govern noncompetes. Binding section 15.50(a) enforces a covenant only if ancillary to an otherwise enforceable agreement and reasonable in time, geography, and restrained activity, imposing no greater restraint than necessary to protect goodwill or another business interest. Tex. Bus. & Com. Code §§ 15.05(a), 15.50(a) [VERIFY]. Meridian bears that burden in this personal-services agreement. Id. § 15.51(b) [VERIFY].

    Binding Texas Supreme Court precedent holds that an at-will employer's executory promise can support a noncompete once performed; providing confidential information reasonably related to a protectable interest suffices. Alex Sheshunoff Mgmt. Servs., L.P. v. Johnson, 209 S.W.3d 644 (Tex. 2006) [VERIFY]; Marsh USA Inc. v. Cook, 354 S.W.3d 764 (Tex. 2011) [VERIFY]. Employment that necessarily requires confidential information can imply the promise, although Meridian made it expressly. Mann Frankfort Stein & Lipp Advisors, Inc. v. Fielding, 289 S.W.3d 844 (Tex. 2009) [VERIFY]. Thus, proof Whitfield actually received confidential pricing and roadmap information would likely establish the ancillary agreement.

    The stated restraint is nevertheless excessive. Meridian has no customers outside three states, yet the covenant reaches the nation; it also bars ownership, management, or any service to a competitor, regardless of relation to Whitfield's work or information. Twenty-four months may be defensible if Meridian proves the information remains competitively useful that long, but the supplied facts do not establish that. Meridian's strongest response is that Whitfield joined a direct competitor in Texas in a related technical role and could exploit current pricing and roadmap knowledge. That supports tailored relief, not the covenant's nationwide, any-capacity ban.

    If the threshold ancillary agreement exists, binding section 15.51(c) says the court shall reform overbroad limits and enforce the covenant as reformed. Tex. Bus. & Com. Code § 15.51(c) [VERIFY]. It bars damages for pre-reformation breach and limits relief to an injunction; knowing overreach may expose Meridian to Whitfield's reasonable fees. Id. [VERIFY].

    
    
    V. Conclusion

    Recommend proposing a restriction limited, at most, to Texas, Louisiana, and Oklahoma and further tailored to Whitfield's accounts, competing pricing or product-roadmap duties, and proven information longevity. Before seeking an injunction, confirm actual disclosures, current sensitivity, Whitfield's accounts and new duties, and evidence of threatened use. Litigation risks include narrower judicial relief, no pre-reformation damages, and possible fee shifting. This draft, including every marked authority, requires independent verification and is not verified legal advice.

MEMORANDUM
To: Supervising Partner
From: Senior Associate
Date: August 21, 2026
Re: Whitfield noncompete — Meridian Freight Systems

Correction. The assignment’s premise that Tex. Bus. & Com. Code § 15.05 is “the Texas covenants-not-to-compete statute” is inaccurate. Section 15.05 prohibits contracts in restraint of trade. [VERIFY] The Covenants Not to Compete Act is §§ 15.50–15.52, which applies “[n]otwithstanding Section 15.05.” [VERIFY]

I. Question Presented
Is a 24-month, nationwide noncompete, signed on day one by at-will Senior Solutions Engineer Dana Whitfield in exchange for a promise of confidential information, enforceable under Texas law, and if not as written, must a court reform it rather than void it?

II. Brief Answer
Not as written, but likely reformable if Meridian furnished confidential information, because the nationwide, any-capacity bar exceeds what is necessary to protect a three-state customer base. If the ancillary-agreement requirement is met, § 15.51(c) requires reformation rather than voiding; a reformed injunction limited to Texas, Louisiana, and Oklahoma and to comparable solutions work would likely still reach her Austin role. [VERIFY]

III. Facts
Houston-based Meridian hired Whitfield on March 3, 2023 as an at-will Senior Solutions Engineer. On her first day, before work, she signed an Employment Agreement containing a 24-month covenant not to “engage in, own, manage, or provide services to any business that competes with the Company anywhere in the United States.” Meridian promised access to confidential information, including customer lists, pricing models, and product roadmaps, without specifying when. Her duties required configuring pricing models for named accounts and attending internal product-roadmap sessions. She resigned June 12, 2026 and accepted a Solutions Architect role at a competing logistics-software company in Austin. Meridian’s customers are in Texas, Louisiana, and Oklahoma; it has never had a customer outside those states.

Facts still needed. Proof of confidential information actually provided; her accounts; Austin overlap; remedy sought; knowledge of overbreadth at signing.

IV. Discussion
Likely unenforceable as written, but reformable and likely to enjoin the Austin job.

Rule. Binding: a covenant is enforceable if ancillary to or part of an otherwise enforceable agreement at the time the agreement is made, with time, geographic, and activity limits that are reasonable and no greater than necessary to protect the promisee’s goodwill or other business interest. Tex. Bus. & Com. Code § 15.50(a). [VERIFY] Meridian bears the burden. Id. § 15.51(b). [VERIFY]

Explanation. At-will employment alone is not an otherwise enforceable agreement. Light v. Centel Cellular Co. of Tex., 883 S.W.2d 642 (Tex. 1994) (binding). [VERIFY] Later provision of promised confidential information can form a unilateral contract, and an implied promise arises if the job reasonably requires that information. Alex Sheshunoff Mgmt. Servs., L.P. v. Johnson, 209 S.W.3d 644 (Tex. 2006) (binding) [VERIFY]; Mann Frankfort Stein & Lipp Advisors, Inc. v. Fielding, 289 S.W.3d 844 (Tex. 2009) (binding). [VERIFY] If the covenant is ancillary but limitations are unreasonable, “the court shall reform the covenant.” No damages for pre-reformation breach; injunctive relief only. Id. § 15.51(c). [VERIFY] Research needed: any authority reforming a non-ancillary covenant.

Application. The express promise plus her duties over three-plus years likely satisfy Sheshunoff and Fielding if the information was provided. A nationwide bar with a three-state customer base is almost certainly greater than necessary. Butler v. Arrow Mirror & Glass, Inc., 51 S.W.3d 787 (Tex. App.—Houston [1st Dist.] 2001, no pet.) (persuasive). [VERIFY] The any-capacity industry restraint is also overbroad. Peat Marwick Main & Co. v. Haass, 818 S.W.2d 381 (Tex. 1991) (binding). [VERIFY]

Counterargument. Whitfield will argue the promise was illusory at signing and that software competes nationally. Sheshunoff and Fielding answer the first point if information was provided. The second is the strongest argument against us, but a three-state customer base is unlikely to support a 50-state bar. Austin is inside that footprint, so reformation would probably still restrain this job. Enforcing the writing as is creates § 15.51(c) fee-shifting risk. [VERIFY]

V. Conclusion
Do not sue to enforce the covenant as written. If we proceed, plead reformation to Texas, Louisiana, and Oklahoma (or customers she served) and to competing solutions work rather than any capacity, and seek injunctive relief only. Open risks: § 15.51(c) fees; no damages after reformation; residual risk of a no-agreement finding.

Not verified legal advice.

MEMORANDUM

To: Supervising Partner
From: Senior Associate
Date: August 21, 2026
Re: Meridian/Whitfield noncompete

Premise note: The assignment's § 15.05 is an antitrust section; the Covenants Not to Compete Act, Tex. Bus. & Com. Code §§ 15.50–.52 [VERIFY], governs. Preliminary; not verified legal advice.

I. Question Presented

Under §§ 15.50–.52, is a 24-month, nationwide, any-capacity noncompete enforceable against an at-will employee who signed it on day one for a promise of confidential information she later received, when the employer's customers are only in Texas, Louisiana, and Oklahoma, and if not, must a court reform it?

II. Brief Answer

Probably enforceable only as reformed. Meridian's promise and delivery of confidential information likely make the covenant ancillary to an otherwise enforceable agreement despite at-will status, but the nationwide, any-capacity scope far exceeds what a three-state business needs. Section 15.51(c) therefore requires reformation, limited to injunctive relief without pre-reformation damages.

III. Facts

Meridian hired Whitfield on March 3, 2023 as a Senior Solutions Engineer; she was at-will throughout. Before starting work that day, she signed an Employment Agreement barring her for 24 months post-separation from engaging in, owning, managing, or providing services to any competing business "anywhere in the United States." The agreement promised her access to confidential information (customer lists, pricing models, product roadmaps), timing unspecified. Her duties required configuring pricing models for named accounts and attending internal roadmap sessions. She resigned June 12, 2026 and accepted a Solutions Architect role with an Austin-headquartered competitor. Meridian has only ever had customers in Texas, Louisiana, and Oklahoma.

Facts still needed: whether Whitfield gave a nondisclosure covenant; her new duties and customer overlap; retained Meridian materials; other consideration.

IV. Discussion

Rule. A noncompete is enforceable if (1) ancillary to or part of an otherwise enforceable agreement when made and (2) its time, geography, and activity limits are reasonable and no broader than necessary to protect the promisee's goodwill or business interest. § 15.50(a) [VERIFY]. Meridian bears the burden. § 15.51(b) [VERIFY].

Explanation. At-will employment is illusory consideration. Light v. Centel Cellular Co., 883 S.W.2d 642 (Tex. 1994) [VERIFY] (binding). But an employer's promise of confidential information supports the covenant once performed. Alex Sheshunoff Mgmt. Servs., L.P. v. Johnson, 209 S.W.3d 644 (Tex. 2006) [VERIFY] (binding); see Marsh USA Inc. v. Cook, 354 S.W.3d 764 (Tex. 2011) [VERIFY] (binding; "reasonably related" consideration suffices). Restraints reaching clients or territory the employee never served, or barring any capacity with any competitor, are overbroad. Peat Marwick Main & Co. v. Haass, 818 S.W.2d 381 (Tex. 1991) [VERIFY] (binding); John R. Ray & Sons, Inc. v. Stroman, 923 S.W.2d 80 (Tex. App.—Houston [14th Dist.] 1996, writ denied) [VERIFY] (binds Harris County trial courts). Two-year terms are commonly upheld [research needed: 24-month Houston cite].

Application. Element one is likely met: Meridian expressly promised confidential information, and Whitfield's pricing and roadmap work shows delivery; under Sheshunoff the covenant became enforceable upon performance despite silence on timing. Element two fails as written: a nationwide ban is unrelated to three-state operations, and "any business that competes" in any capacity is an industry-wide exclusion. Twenty-four months is likely reasonable.

Reformation. If element one is met but limits fail, the court "shall reform" and enforce as reformed, awarding only injunctive relief and no pre-reformation damages; if Meridian knowingly sues on an overbroad covenant as written, it may owe Whitfield's fees. § 15.51(c) [VERIFY]. One Houston court declined to reform at the temporary-injunction stage. Cardinal Health Staffing Network, Inc. v. Bowen, 106 S.W.3d 230 (Tex. App.—Houston [1st Dist.] 2003, no pet.) [VERIFY] (binds Harris County trial courts; research needed on split).

Counterargument. Whitfield's best argument: absent her nondisclosure promise, there is no "otherwise enforceable agreement," only illusory at-will promises, so the covenant is void and unreformable. Secondarily, Meridian's drafting shows knowing overbreadth, inviting fee-shifting.

Conclusion. A covenant reformed to Texas, Louisiana, and Oklahoma and to her actual work is likely enforceable against her Austin role.

V. Conclusion

Recommended: (1) obtain the full agreement; confirm a nondisclosure clause; (2) investigate her new duties and retained data; (3) if proceeding, plead only reformed, injunctive relief to limit § 15.51(c) fee exposure; (4) weigh a Texas-only temporary injunction given reformation-timing risk. Risks: no pre-reformation damages; fee-shifting; missing facts.

# MEMORANDUM

To:    Partner, The Richter Firm, PLLC  
From:  [Associate Name]  
Date:  August 21, 2026  
Re:    Enforceability and potential reformation of Meridian Freight Systems, Inc.’s noncompetition covenant against Dana Whitfield under Texas law

## I. Question Presented
Whether a 24-month nationwide noncompetition covenant signed by an at-will employee on her first day, where the employer promised access to confidential information and the employee’s role involved pricing models and product roadmap sessions, is enforceable under Texas law, and if not, whether a Texas court may reform it rather than void it.

## II. Brief Answer
The covenant is likely enforceable in part but overbroad in scope of territory and potentially activities and duration. Texas law enforces noncompetes ancillary to otherwise enforceable agreements that provide confidential information and reasonably limited in time, geography, and scope; here the promise and apparent provision of confidential information support consideration, but a nationwide restriction exceeds Meridian’s operational footprint. A court is likely to reform the covenant to a reasonable geographic scope aligned with Meridian’s markets, and potentially to refine activity restrictions and duration, rather than void it. Attorney’s fees to the employer on reformation are generally unavailable absent willful breach post-reformation notice. [VERIFY]

## III. Facts
Meridian Freight Systems, Inc., a Houston-based logistics software company, hired Dana Whitfield as a Senior Solutions Engineer on March 3, 2023; her employment was at-will. On her first day, before beginning work, she signed an Employment Agreement containing a covenant not to compete. The covenant prohibits Whitfield, for 24 months after separation, from engaging in, owning, managing, or providing services to any business that competes with the Company anywhere in the United States. In the same agreement, Meridian promised to provide her access to confidential and proprietary information, including customer lists, pricing models, and product roadmaps; the agreement does not specify when access would be provided. Her duties required her to configure pricing models for named accounts and attend internal product roadmap sessions. She resigned on June 12, 2026 and accepted a Solutions Architect role at a competing logistics software company headquartered in Austin, Texas. Meridian’s customers are located in Texas, Louisiana, and Oklahoma; Meridian has never had a customer outside those three states.

### Facts still needed
Whether and when Meridian actually provided Whitfield the promised confidential information, the extent and continuity of such access, any confidentiality and non-disclosure obligations and compliance, the nature of her post-resignation duties and customer contact at the competitor, whether Meridian sells nationally or markets beyond current customers, whether consideration beyond access to information was provided, and whether any non-solicitation or non-disclosure covenants exist.

## IV. Discussion
### Conclusion
The covenant is likely ancillary to an otherwise enforceable agreement and supported by confidential-information consideration, but its nationwide scope and 24-month duration likely render it overbroad. A Texas court would likely reform the covenant to the three-state market (and possibly narrow activities/duration) rather than void it. [VERIFY]

### Rule
Under Texas law, a noncompetition covenant is enforceable if it is ancillary to or part of an otherwise enforceable agreement at the time the agreement is made and contains reasonable limitations as to time, geographical area, and scope of activity that are no greater than necessary to protect the employer’s goodwill or other business interests. If a covenant is overbroad, courts must reform it to the extent necessary and may then enforce it as reformed; damages for pre-reformation breach are limited, and fee-shifting to the employer for obtaining reformation is restricted. Tex. Bus. & Com. Code §§ 15.50–15.52. [VERIFY]

### Explanation with authority
An employer’s promise to provide confidential information can furnish the necessary consideration; access to proprietary customer lists, pricing models, and product plans typically qualifies if actually provided. Reasonableness is assessed in light of the employer’s business, the employee’s role, and the competitive interest to be protected; geographic limits should generally align with the territory where the employee worked or the employer conducts business, and activity restraints should be tailored to the employee’s duties. Overbroad covenants are not void ab initio; Texas courts reform them to reasonable limits and enforce as reformed, subject to statutory limits on damages and fees. Tex. Bus. & Com. Code §§ 15.50–15.52 (binding). [VERIFY]

### Application
Ancillary agreement/consideration: Meridian promised to provide confidential information (customer lists, pricing models, product roadmaps). Whitfield’s role required her to configure pricing models and attend internal roadmap sessions, suggesting she received or necessarily would receive such information. This supports an otherwise enforceable agreement at inception. Risk: if evidence shows the information was not actually provided or not confidential, ancillary consideration could be challenged; factual development is needed.

Scope—geography: Meridian has customers only in Texas, Louisiana, and Oklahoma and has never had a customer elsewhere. A nationwide restraint is broader than necessary to protect Meridian’s goodwill and confidential information. A court would likely narrow the territory to the areas where Meridian does business or where Whitfield had material customer contact, likely the three-state footprint. [VERIFY]

Scope—activities: The covenant bars Whitfield from engaging in, owning, managing, or providing services to any competing business in any capacity. This is broader than necessary to protect Meridian’s interests, as it could preclude non-competitive roles at a competitor. A court is likely to limit the restraint to roles involving competitive activities similar to Whitfield’s work (e.g., pricing configuration, solution engineering tied to Meridian’s customers or products). [VERIFY]

Duration: Twenty-four months is on the longer side but may be upheld depending on industry and sensitivity of information; Texas courts have enforced one- to two-year periods when otherwise reasonable. Given the overbroad territory and activity scope, a court could either uphold 24 months once narrowed or reduce duration to a shorter reasonable period. [VERIFY]

Reformation: If overbroad, the statute authorizes and directs reformation to reasonable limits and enforcement as reformed. Fee-shifting and damages for pre-reformation breach are constrained, which may affect remedies but supports the likelihood of equitable narrowing rather than voiding. Tex. Bus. & Com. Code § 15.51. [VERIFY]

### Strongest counterargument
Meridian may argue nationwide scope is reasonable because logistics software markets and confidential product roadmaps have national relevance and Whitfield joined a competitor headquartered in Austin serving nationwide clients. Conversely, Whitfield may argue lack of specific timing on the promised access and the breadth of “any capacity” render the covenant non-ancillary and unreasonable, and that Meridian’s lack of customers outside three states undermines any national restraint. On a thin record, uncertainty about actual provision of confidential information and Whitfield’s post-resignation duties could cut against enforcement. [VERIFY]

### Conclusion
A Texas court is likely to find the covenant supported by consideration but overbroad in territory and activity, and thus to reform it to a three-state region and limit activities to competitive functions similar to Whitfield’s role, with the 24-month duration either maintained or modestly reduced. [VERIFY]

## V. Conclusion
Meridian has a colorable basis to enforce a reformed covenant. Recommended next steps: gather evidence of Whitfield’s actual access to and use of confidential customer lists, pricing models, and roadmaps; document Meridian’s sales footprint and Whitfield’s customer contacts; assess Whitfield’s current duties at the competitor; and prepare for likely reformation to Texas, Louisiana, and Oklahoma with activity-based limitations and a one- to two-year term. Open risks include potential findings that confidential information was not provided or remains stale, and limitations on pre-reformation damages and attorney’s fees. Research needed: precise Texas authority on reasonableness of nationwide restraints for software companies with regional customers; parameters for reforming “any capacity” role bans; availability of fees and damages post-reformation under § 15.51.

---

**CoCounsel's Draft skill uses a combination of the underlying large language model (LLM), your provided materials, and, when explicitly cited, Thomson Reuters proprietary content to generate legal and non-legal documents. Any citations, including hyperlinked citations to Thomson Reuters content, should be independently verified. As a legal professional, you are ultimately responsible for ensuring the final work is accurate and does not misrepresent the law.**