🔍 Read the full analysis: Korean Judges Reflect On Fake Precedents In An AI-Driven Legal Landscape on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
South Korea’s National Court Administration published a survey on November 2, 2026, in which judges reported encountering nonexistent or distorted legal citations in court filings, including material linked to AI use. The results describe judges’ experiences, not the share of all filings affected; the administration says it plans to propose procedural-law changes allowing fines.
South Korea’s National Court Administration said on November 2 that judges have encountered nonexistent or distorted legal citations in court filings, including fabricated precedents associated with AI use. In a survey of 360 judges, about 80% of responding judges handling civil, family and administrative cases said they had seen such citations, prompting calls for rules including fines and citation-checking duties.
The survey was conducted from October 6 to 16 among judges at courts of all levels. The administration’s release said false citations can impose an unnecessary verification burden on opposing parties and hinder the speed of trials. It did not report how many judges were invited to take part or provide the precise wording of the questions.
Among judges who reported encountering false material, 196 (80.7%) identified precedents that did not exist, while 159 (65.4%) cited real precedents presented with false content. Multiple answers were allowed. Other reported examples included nonexistent statutes and fabricated or nonexistent academic sources.
The survey also asked who submitted the material. 164 judges (69.5%) named the litigants themselves, and 136 (57.6%) named lawyers acting as representatives or defense counsel. The administration said false citations appeared in documents filed by both parties and lawyers. Those responses reflect judges’ reports, not a count of confirmed incidents or a finding of misconduct in individual cases.
Pressure for Rules on Court Filings
The findings put attention on a practical risk for courts: judges and opposing parties may need to spend time checking citations that look authentic but cannot be verified. The administration’s release said this burden can slow proceedings. The survey does not establish how often this occurs across all filings, but it records a substantial number of judges saying they have encountered the problem.
Judges overwhelmingly favored an institutional response: 91.7% said one was needed, and 64.8% called it “very necessary.” The most frequently selected options were a legal basis for fines, a duty to verify cited precedents and a duty to disclose AI use. These are reported preferences, not rules already in force.
The issue reaches beyond generative AI. A fabricated citation can enter a filing through an AI tool, a mistaken manual citation or other means. The survey describes examples that include AI-generated judgments in the format of Supreme Court decisions, real rulings cited inaccurately and chatbot conversation text or litigation-strategy notes pasted into documents. Each can require the court to distinguish a valid authority from material that only appears credible.
Products Worth Considering
Amazon Kindle Paperwhite Signature Edition, 32GB (2024), Metallic Black
Anker SOLIX C300 Portable Power Station, 288Wh
DJI Osmo Mobile 8 3-Axis Gimbal Stabilizer for Phones
As an affiliate, we earn on qualifying purchases.
What the Survey Actually Measures
The administration’s release described the survey as a response to concern that AI hallucinations can produce false or nonexistent precedent citations. Its figures measure what judges say they have encountered; they do not measure the proportion of court filings containing fabricated material or prove that AI generated every example.
The widely reported “about 80%” figure applies to responding judges handling civil, family and administrative cases. It should not be read as 80% of all 360 survey respondents. Kyunghyang Shinmun reported that 243 of the 360 judges (67.5%) said they had encountered fake citations. The administration’s release did not print that overall figure; the reported count is consistent with the denominator behind the specific citation-type percentages.
A translation issue also illustrates why checking the underlying source matters. An English-language news version republished by The Korea Herald described 164 judges as saying they themselves had cited false precedents. The Korean release instead identifies the parties themselves as a source of the citations. The survey does not say those judges admitted filing false material.
Limits of the Judges’ Responses
The survey does not say how many judges were invited, how representative the 360 respondents are, or how frequently false citations appear in filings overall. It records judges’ reported experience rather than independently verified case counts. The results also do not establish whether each fabricated or distorted citation was produced by AI, knowingly submitted or the product of an ordinary error.
The administration’s release outlines possible procedural safeguards, but the source material does not provide a finalized bill, a timetable for submitting it or details on how any fine or verification duty would work. The proposed requirements and their scope remain unsettled.
Proposed Changes to Procedure Laws
The National Court Administration plans to propose changes to South Korea’s Civil and Criminal Procedure Acts that would provide a legal basis for fines, according to the source material. The surveyed judges also favored citation-verification and AI-disclosure duties, though the administration has not, in the material provided, set out final legislative language or said when proposals will be submitted.
The next developments to watch are whether the administration formally advances those changes and how lawmakers define the conduct that could trigger a penalty. Until then, the survey offers evidence of judges’ reported concerns, not a new legal obligation for parties or lawyers.
Key Questions
What did South Korea’s judges report?
In a survey published November 2, about 80% of responding civil, family and administrative judges said they had encountered nonexistent or distorted precedents. The survey covered 360 judges across courts of all levels.
Does the 80% figure refer to all 360 judges?
No. It refers to responding judges handling civil, family and administrative cases. Kyunghyang Shinmun reported that 243 of all 360 respondents (67.5%) had encountered fake citations.
Did the survey prove that AI created all the fake citations?
No. The administration cited AI hallucinations as a concern, and the examples included AI-generated material, but the survey does not establish the source of every false citation.
What safeguards did judges support?
91.7% said an institutional response was needed. Popular options included a legal basis for fines, a duty to verify citations and a duty to disclose AI use. These are preferences reported in the survey, not current requirements.
What happens next?
The administration plans to propose changes to the Civil and Criminal Procedure Acts that would allow fines. The timing, details and prospects for passage are not specified in the source material.
Source: ThorstenMeyerAI.com
