A 360-Judge Survey Tackles AI And Fake Precedents In Korean Courts
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: A 360-Judge Survey Tackles AI And Fake Precedents In Korean Courts on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

South Korea’s National Court Administration published a survey of 360 judges who responded between October 6 and 16, 2026. About 80% of responding civil, family and administrative judges said they had encountered false or nonexistent precedents; the survey measures judges’ reported experience, not the share of court filings containing fakes.

South Korea’s National Court Administration said a survey of 360 judges found that about 80% of responding judges handling civil, family and administrative cases had encountered false or nonexistent precedents in court filings. The November 2 results describe judges’ reported experiences—not the proportion of filings containing fabricated citations—and the administration says it plans to propose procedural-law changes that could allow fines.

The survey was conducted from October 6 to 16, 2026, among judges at courts of all levels. The court administration’s release says AI hallucinations can produce false case citations, creating extra verification work for opposing parties and potentially slowing proceedings. It does not state how many judges were invited to participate or provide the survey questions.

Among judges who reported encountering false citations, 196 (80.7%) identified precedents that did not exist, while 159 (65.4%) identified real precedents cited with false content. Respondents could select more than one type. The release also lists false or nonexistent statutes, reported by 41 judges, and papers or other literature, reported by 10.

The denominator matters: the court administration’s “about 80%” finding concerns responding judges in civil, family and administrative cases who answered the relevant question. Kyunghyang Shinmun reported that 243 of all 360 respondents (67.5%) had encountered fake citations. That broader figure is not printed in the administration’s release. Judges also said the material appeared in filings by litigants themselves and by lawyers: 164 (69.5%) named parties, and 136 (57.6%) named lawyers acting as representatives or defense counsel.

At a glance
reportWhen: Survey conducted October 6–16, 2026; re…
The developmentSouth Korea’s National Court Administration released its first survey of judges’ experiences with fake and distorted citations in court filings.
Video: Fake AI Precedents in Korean Courts: What a Survey of 360 Judges Found. Narration: AI voice clone of Thorsten Meyer; images AI-generated.

Pressure for Rules on Court Filings

The survey puts a number on a practical concern for courts: judges say they are spending time checking whether cited legal authorities exist and say that false citations can burden opposing parties and impede timely trials. The reported examples range from invented Supreme Court decisions to genuine rulings presented with altered or misleading reasoning. This is not only an issue of AI use; the findings also identify filings by both self-represented parties and licensed lawyers.

91.7% of respondents called for an institutional response. The most selected options were a legal basis for fines (61.4%), a duty to verify citations (61.1%), and a duty to disclose AI use (60.6%). Those figures represent judges’ preferred measures in the survey, not rules already adopted or evidence that any one measure would prevent fabricated citations.

The proposed direction could affect how lawyers and litigants prepare submissions and what courts do when a citation proves false. Fines would mark a shift from relying on professional care and judicial checking toward an explicit procedural penalty. The details—including who would be subject to penalties and how intent or error would be assessed—have not yet been set out in the material provided.

Products Worth Considering

Apple Watch SE (2nd Gen) GPS 40mm with Starlight Aluminum Case and Sport Band (S/M)

Apple Watch SE (2nd Gen) GPS 40mm with Starlight Aluminum Case and Sport Band (S/M)

DJI Mic 3 Wireless Microphone Kit (2 Transmitters, 1 Receiver, Charging Case)

DJI Mic 3 Wireless Microphone Kit (2 Transmitters, 1 Receiver, Charging Case)

Apple Watch Series 8 GPS, 41mm, Midnight Aluminum Case with Midnight Sport Band (S/M)

Apple Watch Series 8 GPS, 41mm, Midnight Aluminum Case with Midnight Sport Band (S/M)

As an affiliate, we earn on qualifying purchases.

What Judges Reported Seeing

The administration’s press release, issued by its Judicial Support Office, describes several forms of false material. These include AI-generated judgments styled to resemble Supreme Court decisions; Supreme Court case numbers paired with reasoning that cannot be found in the cited ruling; real decisions whose holdings have been changed or misrepresented; and text from AI conversations or litigation-strategy notes pasted into a filing without editing.

The release says false citations were found in documents submitted by both parties and lawyers. A translation discrepancy illustrates why attribution and source-checking matter here: an English-language Yonhap report republished by The Korea Herald said 164 judges had cited false precedents themselves. The Korean release instead says those judges identified the parties themselves as a source of false material. The survey does not say that judges admitted to making false citations.

The court administration’s stated concern is that fabricated authorities shift verification work to opposing parties and can hinder the speed of trials. The survey is a report of judicial experience, not an audit of filings, and its results do not establish how often the problem occurs across all Korean cases.

Limits of the Survey Evidence

The results do not establish the rate of fabricated citations in Korean court filings. The survey records what judges said they had encountered; the release does not provide a filing-level review, the number of judges invited, or the exact wording of its questions. The headline figure also uses a narrower group of civil, family and administrative judges, while the 67.5% figure across all respondents was reported by Kyunghyang Shinmun rather than printed in the release.

It is not clear how the proposed duties or fines would be defined, which court filings they would cover, or what process would determine whether a false citation merits a penalty. The source material describes plans to propose amendments, not enacted legislation. It also does not quantify how many reported instances involved AI rather than other sources of error or fabrication.

Proposed Procedure Law Changes

The National Court Administration plans to propose amendments to South Korea’s Civil and Criminal Procedure Acts that would provide a legal basis for fines, according to the source material. The survey also found support for citation-verification and AI-disclosure duties, but the material does not say whether those measures will be included in the proposals or when any bill will be presented.

The next concrete developments will be the text and timing of any proposed amendments, followed by consideration through the legislative process. Until those details emerge, the survey documents judges’ concerns and preferences; it does not change current filing rules or establish new penalties.

Key Questions

What did the Korean judges’ survey find?

The National Court Administration said about 80% of responding civil, family and administrative judges had encountered false or nonexistent precedents. Across all 360 respondents, Kyunghyang Shinmun reported that 243, or 67.5%, had seen fake citations.

Does the survey show that 80% of court filings contain fake citations?

No. The figure describes judges’ reported experience, not the percentage of filings containing fabricated material. The release does not include a filing-level audit or estimate.

Were lawyers implicated in the reported cases?

Yes. The survey says 136 judges, or 57.6% of those answering that question, named lawyers acting as representatives or defense counsel as a source of false material. Respondents could select more than one source.

What changes does the court administration plan to seek?

The administration plans to propose amendments to the Civil and Criminal Procedure Acts to allow fines. Judges also supported a duty to verify citations and a duty to disclose AI use, but the source does not confirm whether those duties will be part of the proposed amendments.

Source: ThorstenMeyerAI.com

You May Also Like

Data Center Surges In Global Coverage

Data center mentions worldwide have increased dramatically, with GDELT reporting 41 times the usual coverage, highlighting growing global interest and activity.

Anthropic’s Nov. 12 Policy Bans Cruelty Toward Its Claude AI

A Yahoo Finance headline says Anthropic will ban cruelty toward Claude from Nov. 12, but the policy’s wording, year and enforcement are unclear.

AI And The Rise Of Invisible Watermarks: What You Need To Know

Anthropic plans to add invisible watermarks to Claude-generated text, aiming to improve AI content detection. Details on implementation and timing remain unclear.

How The Railway Breach Exposed Demand For AI Keys

Railway says an attacker may have accessed some customer environment variables and links the incident to attacks targeting AI inference providers.