CONTENTS
- 1. Violation of the Attorney-at-Law Act | Overview of the Subject Case
- 2. Violation of the Attorney-at-Law Act | Problems with Citing AI-Generated Fictitious Precedents
- - Similar Cases Identified in Korea
- - Overseas Cases and the Response in Various Countries
- 3. Violation of the Attorney-at-Law Act | Matters That May Be Problematic Under the Korean Legal System
- - The Significance of the Case and Future Tasks for the Legal Profession
1. Violation of the Attorney-at-Law Act | Overview of the Subject Case
A violation of the Attorney-at-Law Act can place an attorney's profession and livelihood under serious threat.
The following describes a matter that recently became an issue in a criminal trial.
In that trial, Attorney A cited five judgments in an opinion brief, stating that they supported the defendant's arguments.
When the court checked those precedents through its electronic database, however, it found that none of them existed and that all were fictitious judgments.
During the proceedings, the attorney admitted to having "used AI," and the panel requested confirmation of the source of the precedents once more.
This case involved the submission of AI-generated, fictitious legal reasoning without any verification of the underlying facts, and it raised questions about possible deception of the court and a breach of an attorney's professional ethics.
2. Violation of the Attorney-at-Law Act | Problems with Citing AI-Generated Fictitious Precedents
Generative AI is highly capable of composing plausible sentences, but it carries the problem of fabricating nonexistent judgments, scholarly materials, and legal reasoning as if they were real.
In this case, Attorney A mistook precedents generated by AI for actual judgments and submitted them, and such conduct gives rise to the following problems.
- Presenting fictitious legal reasoning as a basis for the court's determination undermines the accuracy and fairness of judicial proceedings
- The attorney's failure to fulfill the duty to verify the facts harms the client's interests
- From the court's standpoint, trial efficiency decreases and the burden of reviewing briefs increases
- At a societal level, trust in AI declines and the quality of legal services deteriorates
Similar Cases Identified in Korea
This is not a problem confined to attorneys alone.
An issue also arose at an investigative agency such as the police, when a complainant who received a non-referral decision prepared with the aid of AI pointed out errors in it.
- In a child abuse complaint case, it was confirmed that the non-referral decision sent by the police cited nonexistent Supreme Court and District Court precedents
- The wording did not appear anywhere in any judgment, and it is presumed that no separate verification was carried out during the drafting process
- Experts pointed out the possibility that investigative agencies as well may have omitted necessary verification steps when using AI-generated information
This case shows that errors in AI-generated information can affect the determination of actual cases not only in the court's adjudication but also at the investigation stage.
Overseas Cases and the Response in Various Countries
Abroad, strong sanctions have already followed in cases involving the submission of AI-generated fictitious precedents.
1) United States
- 2025, Utah Court of Appeals: ordered an attorney who submitted fictitious precedents to make a $1,000 donation
- Indiana federal court: imposed a $6,000 fine on an attorney who submitted fictitious materials
- American Bar Association (ABA): issued a formal opinion that reviewing AI outputs is part of an attorney's "duty of competent representation"
- Between 2023 and 2025, 95 cases of submitting AI-generated fictitious materials were confirmed (58 cases as of 2025)
2) United Kingdom
- In a case involving an attorney who submitted five nonexistent precedents during administrative litigation, the High Court warned of a "possibility of contempt of court"
- If contempt of court is found, a fine or up to two years' imprisonment is possible
In the United Kingdom, there is a strong view that, regardless of whether an attorney used AI, the submission of false statements itself constitutes deception of the court.
3. Violation of the Attorney-at-Law Act | Matters That May Be Problematic Under the Korean Legal System
An attorney who is found to have cited AI-generated fictitious precedents as they were risks violating the following provisions.
1) Article 24 of the Attorney-at-Law Act
"An attorney shall not conceal the truth or make false statements"
"An attorney shall not engage in conduct that damages dignity"
2) Code of Ethics for Attorneys
Article 5 (Maintenance of Dignity) : "An attorney shall maintain dignity and shall not engage in conduct that damages honor"
Article 36 (Trial Procedure) : "An attorney shall not intentionally assert false facts or submit false evidence"
Even though this case did not involve an intentional submission, the failure to fulfill the basic duty to verify the facts may lead to questions about a breach of an attorney's professional ethics.
The Significance of the Case and Future Tasks for the Legal Profession
This case offers an early indication of the new risks the legal profession faces as the use of AI rapidly spreads.
In particular, it leaves the following implications.
- The reliability of precedents that AI provides is not absolute, and all such information requires final verification by the attorney
- Structural factors, such as the nondisclosure of judgments and limits on searching lower-court precedents, cannot be ignored, so institutional improvements will be needed going forward
- AI can instead increase the time spent reviewing briefs, giving rise to problems of reduced trial efficiency
- The need to establish guidelines on the use of AI at the level of the courts and the bar association is coming to the fore
AI can serve as a tool to improve the efficiency of legal services, but if legal professionals use it uncritically, it can instead become a risk factor that may undermine confidence in the judiciary and the rights of clients.
Daeryun Law Firm LLP is advancing legal technology, including the adoption of AI and platforms, to enhance client convenience and to lead innovation in legal services.
Recognizing deeply the heavy responsibility for fact verification and personal data protection that technological development brings, Daeryun will make consistent ethical awareness the foundation of its services.
Through Daeryun's carefully designed one-stop legal service that reconciles innovation and trust, we are committed to protecting our clients' rights as securely as possible.
