Live tracker

SlopJudges

Tracking AI slop in real U.S. Supreme Court, Courts of Appeals, and federal District Court opinions. Every opinion is pulled from the public CourtListener record and scored by our own blend of an LLM judge and deterministic heuristics — no third-party detector.

Opinions analyzed
3,128
Slop opinions
98
Courts tracked
26
Last updated
Nov 15, 2026, 2:42 PM

Cumulative AI-slop opinions over time

Supreme Court vs. the regional Courts of Appeals vs. the specialized D.C. & Federal Circuits vs. the federal District Courts.

02040608020192020202120222023202420252026

Justices by AI-slop opinions

Supreme Court justices ranked by opinions flagged as slop; ties break on average AI score.

  1. 1
    Barrett
    11 avg AI · 8 opinions
    AI 20
    1
    ↳ Watson v. Republican National Committee
  2. 2
    Kavanaugh
    9 avg AI · 24 opinions
    AI 18
    1
    ↳ West Virginia v. B. P. J. (Kavanaugh, Majority)
  3. 3
    Kagan
    9 avg AI · 18 opinions
    AI 16
    0
    ↳ Mullin v. Doe (Kagan, dissenting)
  4. 4
    Jackson
    8 avg AI · 40 opinions
    AI 17
    0
    ↳ Trump v. Cook (Jackson, concurring)
  5. 5
    Roberts
    8 avg AI · 10 opinions
    AI 15
    0
    ↳ Trump v. Barbara (Roberts, Majority)
  6. 6
    Sotomayor
    8 avg AI · 24 opinions
    AI 11
    0
    ↳ Trump v. California (Sotomayor, dissenting)
  7. 8
    Gorsuch
    7 avg AI · 34 opinions
    AI 14
    0
    ↳ Trump v. Barbara (Gorsuch, dissenting)
  8. 9
    Thomas
    7 avg AI · 33 opinions
    AI 14
    0
    ↳ Bondi v. Vanderstok (Thomas, dissenting)
  9. 10
    Alito
    7 avg AI · 23 opinions
    AI 10
    0
    ↳ Chatrie v. United States (Alito, dissenting)
  10. 11
    Breyer
    6 avg AI · 6 opinions
    AI 7
    0
    ↳ Johnson v. Guzman Chavez (Breyer, dissenting)
  11. 12
    Per Curiam
    4 avg AI · 14 opinions
    AI 12
    0
    ↳ NVIDIA v. E. Ohman J:or Fonder AB
  12. 13
    Ketanji Brown Jackson
    3 avg AI · 2 opinions
    AI 4
    0
    ↳ United States v. Miller
  13. 14
    John G. Roberts
    2 avg AI · 3 opinions
    AI 3
    0
    ↳ Loper Bright Enterprises v. Raimondo
  14. 15
    Sonia Sotomayor
    2 avg AI · 2 opinions
    AI 2
    0
    ↳ Wilkinson v. Garland
  15. 16
    Amy Coney Barrett
    1 avg AI · 1 opinion
    AI 1
    0
    ↳ Acheson Hotels, LLC v. Laufer
  16. 17
    Neil Gorsuch
    1 avg AI · 4 opinions
    AI 1
    0
    ↳ Bondi v. Vanderstok
  17. 18
    Clarence Thomas
    0 avg AI · 1 opinion
    AI 0
    0
    ↳ New York State Rifle & Pistol Assn., Inc. v. Bruen

Courts by AI-slop opinions

All tracked courts ranked by opinions flagged as slop; ties break on average AI score.

  1. 1
    D.D.C.
    5 avg AI · 747 opinions
    AI 35
    23
    ↳ United States v. Senior
  2. 2
    7th Cir.
    6 avg AI · 313 opinions
    AI 37
    21
    ↳ American Academy of Pediatrics v. James Uthmeier
  3. 3
    8th Cir.
    6 avg AI · 203 opinions
    AI 31
    14
    ↳ United States v. Alexzandra Blanco
  4. 4
    11th Cir.
    5 avg AI · 112 opinions
    AI 25
    6
    ↳ Annette Kingsolver v. U.S. Attorney General
  5. 6
    9th Cir.
    4 avg AI · 207 opinions
    AI 25
    5
    ↳ United States v. Abdul Kareem
  6. 7
    4th Cir.
    6 avg AI · 176 opinions
    AI 50
    4
    ↳ United Financial Casualty Co. v. Greg Ball
  7. 8
    3rd Cir.
    6 avg AI · 127 opinions
    AI 26
    4
    ↳ United States v. Davone Walker
  8. 9
    5th Cir.
    5 avg AI · 199 opinions
    AI 24
    4
    ↳ Students Engaged in Advancing Texas v. Ken Paxton
  9. 10
    6th Cir.
    5 avg AI · 173 opinions
    AI 32
    3
    ↳ In re: E. Palestine Train Derailment
  10. 11
    SCOTUS
    7 avg AI · 249 opinions
    AI 20
    2
    ↳ Watson v. Republican National Committee
  11. 12
    1st Cir.
    6 avg AI · 152 opinions
    AI 23
    2
    ↳ Rana v. Blanche
  12. 13
    10th Cir.
    5 avg AI · 102 opinions
    AI 24
    2
    ↳ Dunn v. Santa Fe Natural Tobacco Company
  13. 14
    D.C. Cir.
    4 avg AI · 108 opinions
    AI 29
    1
    ↳ Ronda L. Davis v. DC
  14. 16
    S.D.N.Y.
    12 avg AI · 2 opinions
    AI 16
    0
    ↳ Rasberry ex rel. Situated v. Columbia Cnty.
  15. 17
    C.D. Cal.
    11 avg AI · 2 opinions
    AI 12
    0
    ↳ Vieira v. Mentor Worldwide, LLC
  16. 18
    E.D.N.Y.
    11 avg AI · 2 opinions
    AI 14
    0
    ↳ United States v. Ceasar
  17. 20
    E.D. Tex.
    7 avg AI · 2 opinions
    AI 7
    0
    ↳ True Health Diagnostics, LLC v. Azar
  18. 21
    E.D. Va.
    7 avg AI · 2 opinions
    AI 7
    0
    ↳ Elhady v. Kable
  19. 22
    N.D. Cal.
    6 avg AI · 2 opinions
    AI 7
    0
    ↳ E. Bay Sanctuary Covenant v. Barr
  20. 23
    N.D. Tex.
    6 avg AI · 2 opinions
    AI 6
    0
    ↳ Ford v. Freemen
  21. 24
    S.D. Tex.
    6 avg AI · 2 opinions
    AI 7
    0
    ↳ Doe v. Salesforce.Com, Inc.
  22. 25
    S.D. Fla.
    5 avg AI · 2 opinions
    AI 10
    0
    ↳ Compere v. Nusret Miami, LLC
  23. 26
    N.D. Ill.
    0 avg AI · 0 opinions
    AI 0
    0

Most AI-like opinions

The highest-scoring opinions in the dataset. On the real federal record almost none cross the flag line (18); these are simply the current leaders, shown with the signals that moved their score.

  • United Financial Casualty Co. v. Greg Ball
    Court of Appeals for the Fourth Circuit · Mar 30, 2023
    AI 50

    The text consists solely of repeated document metadata with no substantive legal content, indicating it is not a genuine judicial opinion.

    repeated document metadatalack of substantive content
    LLM 100Heuristic 0Confidence 20%
  • Petersen Energ�a; Eton Park v. Argentie Argentine Republic, YPF S.A.
    Court of Appeals for the Second Circuit · Mar 27, 2026
    AI 47

    The text contains malformed citations and potentially fabricated case names, indicating a high likelihood of AI assistance.

    malformed reporter citationsfabricated case names
    LLM 85Heuristic 12Confidence 18%
  • American Academy of Pediatrics v. James Uthmeier
    Court of Appeals for the Seventh Circuit · Sep 14, 2026 · Scudderconcurs and dissents
    AI 37

    Although much of the prose resembles authentic appellate writing, the co-occurrence of serious citation anomalies, likely fabricated future citations, and some polished essay-style structuring makes substantial AI generation likely.

    citation problems and apparent hallucinationswrong Dombrowski citation volume repeatedimpossible cert-denied date preceding appellate decisionsome essay-like framing and tidy Roman-numeral structureLLM-favored phrases such as 'At its core' and broad thematic framing
    LLM 78Heuristic 10Confidence 86%
  • United States v. Senior
    District Court, District of Columbia · Oct 6, 2026 · Judge Colleen Kollar-Kotelly
    AI 35

    Heuristic-only score (LLM judge unavailable).

    Unusually uniform sentence rhythmSparse citation density for opinion length
    LLM 0Heuristic 35Confidence 30%
  • In re: E. Palestine Train Derailment
    Court of Appeals for the Sixth Circuit · Nov 10, 2026 · Amul R. Thapar; Chad A. Readler; Whitney D. Hermandorfer
    AI 32

    The excerpt reads like a polished appellate opinion with concrete record citations, varied punchy and clause-heavy sentences, and no obvious LLM boilerplate or generic essay structure, though the future/unverifiable citations slightly limit confidence.

    Unusually uniform sentence rhythmSparse citation density for opinion lengthprecise and conventionally formatted citationsvaried sentence lengths and judicial rhythmno chatbot boilerplate or listicle scaffolding
    LLM 10Heuristic 37Confidence 82%
  • United States v. Alexzandra Blanco
    Court of Appeals for the Eighth Circuit · Aug 10, 2026
    AI 31

    The text otherwise reads like a conventional Eighth Circuit opinion with concrete facts, standard formatting, precise statutory/rule citations, and no chatbot-style framing or buzzword clusters, but the malformed Campbell citation is a notable isolated AI-like warning sign.

    Unusually uniform sentence rhythmisolated impossible reporter/year citation: 6 F.4th with 2014 datesome repetitive appellate-disposition phrasing
    LLM 22Heuristic 33Confidence 78%
  • Cassandra Sykes v. Experian Information Solutions, Inc.
    Court of Appeals for the Seventh Circuit · Jul 30, 2026 · Jackson-Akiwumi
    AI 31

    The excerpt reads like a conventional appellate opinion with precise and plausible citations, normal judicial structure, varied sentence rhythm, and no chatbot boilerplate, buzzword clustering, listicle scaffolding, or citation anomalies apparent from the text.

    Unusually uniform sentence rhythmChatbot-style holding recap
    LLM 5Heuristic 38Confidence 90%
  • L.M. Caheb v. Mullin
    District Court, District of Columbia · Nov 10, 2026 · Chief Judge James E. Boasberg
    AI 30

    The excerpt is too short and truncated to assess meaningfully, but it shows ordinary court-caption and introductory-opinion language with no obvious generative-AI tells.

    Unusually uniform sentence rhythm
    LLM 5Heuristic 35Confidence 2%
  • Patel v. Atlantic Monthly Group LLC
    District Court, District of Columbia · Nov 10, 2026 · Judge Emmet G. Sullivan
    AI 29

    The excerpt is only a caption and opening sentence with standard judicial formatting and no discernible generative-AI tells, making any assessment highly uncertain.

    Unusually uniform sentence rhythm
    LLM 5Heuristic 34Confidence 2%
  • Ward v. Mayorkas
    District Court, District of Columbia · Sep 11, 2026 · Judge Reggie B. Walton
    AI 29

    Although the prose and structure otherwise resemble a genuine D.D.C. memorandum opinion, the assertion that Markwayne Mullin is the current Secretary of Homeland Security and the related substitution history are blatant factual/citation-style anomalies that strongly suggest synthetic or substantially AI-assisted drafting.

    fabricated or impossible official-capacity defendant statementsuspicious party substitution footnotepossibly hallucinated future/implausible citationotherwise polished opinion-like formatting with precise but questionable citations
    LLM 72Heuristic 0Confidence 82%
  • Rapaport v. Nivoda
    Court of Appeals for the Second Circuit · Sep 4, 2026
    AI 29

    Heuristic-only score (LLM judge unavailable).

    Unusually uniform sentence rhythm
    LLM 0Heuristic 29Confidence 30%
  • Ronda L. Davis v. DC
    Court of Appeals for the D.C. Circuit · Sep 4, 2026
    AI 29

    Heuristic-only score (LLM judge unavailable).

    Unusually uniform sentence rhythm
    LLM 0Heuristic 29Confidence 30%
  • Maria Reynoso-Salgado v. Todd W. Blanche
    Court of Appeals for the Seventh Circuit · Sep 1, 2026 · Jackson-Akiwumi
    AI 27

    The prose largely resembles a genuine appellate opinion, but multiple citation anomalies—especially the impossible-looking pin cites—are strong machine-generation tells that raise the likelihood of substantial AI assistance.

    malformed pinpoint citation “1191–13”internal citation inconsistency for State v. Evans, 492 N.W.2d 141 later cited at 481several post-2024 citations and party names that appear potentially fabricated or difficult to verify from form alonevery tidy categorical-approach structure
    LLM 58Heuristic 7Confidence 78%
  • Liu v. Noem
    District Court, District of Columbia · Jul 29, 2026 · Judge Jia M. Cobb
    AI 27

    The prose largely reads like a conventional judicial opinion, but the highly implausible official-capacity defendant substitution and possibly fabricated future case citations are strong hallucination-style signals that substantially raise the likelihood of AI assistance.

    implausible party substitution naming Markwayne Mullin as Secretary of Homeland Securitypossible fabricated or anachronistic supporting cases with 2026 WL citationsotherwise precise but potentially hallucinated citation scaffoldingno chatbot boilerplate or obvious buzzword clusters
    LLM 68Heuristic 1Confidence 78%
  • Perales v. Office of the Clerk
    District Court, District of Columbia · Jul 23, 2026 · Judge Jia M. Cobb
    AI 27

    The excerpt is a very short, standard dismissal order with no chatbot phrasing, buzzword clusters, citation issues, or structural AI tells, though its brevity limits confidence.

    Unusually uniform sentence rhythm
    LLM 2Heuristic 32Confidence 7%
  • Truesdale v. District of Columbia Government
    District Court, District of Columbia · Oct 23, 2026 · Judge Dabney L. Friedrich
    AI 26

    The order is concise, case-specific, and conventionally judicial with precise docket references, but a notable impossible reporter citation and a stray drafting error create some suspicion without amounting to strong LLM evidence.

    Unusually uniform sentence rhythmSparse citation density for opinion lengthone impossible/malformed citation: F.3d reporter with 1964 dateminor proofreading error: 'under the of professional conduct'otherwise concrete docket references and case-specific reasoning
    LLM 22Heuristic 27Confidence 78%
  • Public Citizen v. United States Department of Justice
    District Court, District of Columbia · Oct 22, 2026 · Judge Trevor N. McFadden
    AI 26

    The text reads like a conventional short federal FOIA order with precise docket references, plausible and properly formatted citations, normal judicial phrasing, and no chatbot boilerplate, buzzword clustering, listicle structure, or obvious citation hallucinations.

    Unusually uniform sentence rhythmChatbot-style holding recap
    LLM 8Heuristic 30Confidence 86%
  • United States v. Davone Walker
    Court of Appeals for the Third Circuit · Sep 15, 2026
    AI 26

    This brief procedural order uses standard appellate formatting and precise docket details with no chatbot phrasing, buzzword filler, mechanical essay structure, or citation anomalies.

    Unusually uniform sentence rhythm
    LLM 2Heuristic 32Confidence 20%
  • United States v. Calvin Carter
    Court of Appeals for the Eighth Circuit · Jul 7, 2026
    AI 26

    The excerpt reads like a conventional concise appellate per curiam opinion with precise Guidelines and case citations, fact-specific reasoning, and no chatbot boilerplate, buzzword clusters, vague legal generalities, or malformed citation patterns suggestive of LLM drafting.

    Unusually uniform sentence rhythmChatbot-style holding recap
    LLM 5Heuristic 32Confidence 93%
  • United States v. Abdul Kareem
    Court of Appeals for the Ninth Circuit · Nov 6, 2026
    AI 25

    Although the format mimics a Ninth Circuit opinion, the text contains several independent machine-generation tells, including a conspicuously wrong rule citation, a likely hallucinated terrorism attribution, and mechanically repetitive opinion-style prose.

    factually suspect attribution of Anwar al-Awlaki as an ISIS leaderapparently wrong appellate rule citation for a criminal appeal: Fed. R. App. P. 4(a)(1)(A)repetitive mechanical factual structure with A-F subsections repeatedly beginning with government-introduced evidencegeneric essay-like opening and tidy framingunusual vague phrasing such as ISIS-related or ISIS-adjacent materials
    LLM 62Heuristic 1Confidence 78%
  • American Academy of Pediatrics v. James Uthmeier
    Court of Appeals for the Seventh Circuit · Sep 14, 2026 · Kolarconcurs
    AI 25

    The prose mostly resembles a genuine appellate opinion with varied rhythm and many precise citations, but the internally implausible and erroneous citations are strong enough hallucination-like signals to put it in a moderate-risk range.

    one clear erroneous citation to Dombrowski as 389 U.S. 479 instead of 380 U.S. 479implausible certiorari citation dated before the lower-court decisionsome essay-like tidy structuring with numbered sections and brief concluding paragraphsoccasional LLM-favored phrasing such as “safety valve” and “underscoring”
    LLM 48Heuristic 10Confidence 74%
  • Annette Kingsolver v. U.S. Attorney General
    Court of Appeals for the Eleventh Circuit · Aug 20, 2026
    AI 25

    The excerpt reads like a conventional appellate opinion with precise, plausible citations, case-specific factual detail, varied judicial sentence structure, and no chatbot framing, buzzword clusters, malformed citations, or essay-like AI scaffolding.

    Unusually uniform sentence rhythm
    LLM 8Heuristic 29Confidence 86%
  • Dunn v. Santa Fe Natural Tobacco Company
    Court of Appeals for the Tenth Circuit · Nov 12, 2026
    AI 24

    The text reads like a conventional appellate order and dissent with precise legal issues, plausible citations, varied sentence structure, and no chatbot boilerplate, buzzword clustering, malformed citations, or generic essay-style filler.

    Unusually uniform sentence rhythm
    LLM 6Heuristic 28Confidence 90%
  • Students Engaged in Advancing Texas v. Ken Paxton
    Court of Appeals for the Fifth Circuit · Oct 28, 2026
    AI 24

    This is a short, formulaic en banc rehearing order with standard court formatting, precise docket references, and no chatbot phrasing, buzzword clusters, mechanical essay structure, or citation irregularities.

    Unusually uniform sentence rhythmHeavy em-dash usage
    LLM 2Heuristic 29Confidence 48%
  • Perez v. v Wax Studio, LLC
    District Court, District of Columbia · Sep 24, 2026 · Judge Rudolph Contreras
    AI 24

    The excerpt is largely plausible law-clerk style, but the cluster of suspect citations, citation-format anomalies, mechanical structure, and small procedural/formatting inconsistencies suggests a meaningful likelihood of substantial generative-AI assistance.

    multiple suspicious or possibly fabricated 2026 WL citationsmalformed citation formatting such as "503 Supp.2d"mechanical judicial template and tidy sectioninginconsistent numbering and capitalization conventionsunusual phrasing such as "For the forthcoming reasons"
    LLM 62Heuristic 0Confidence 74%