What are the consequences of the misuse of AI in legal cases? (update)

The rise of AI technology is having a huge impact on society worldwide, and the judiciary in the Netherlands is not escaping this massive transformation either.

We previously wrote the article below about the consequences of the misuse of AI systems in court cases:

what are the consequences of misuse of ai in court cases

Developments are following one another at a rapid pace. We would like to provide you with an update as of May 2026.

The Dutch Enterprise Chamber also affected

The Dutch Enterprise Chamber is authorized to hear and decide inquiry procedures (‘enquêteprocedures’) and certain rights and powers of works councils. In short, the Enterprise Chamber is a specialized court that deals primarily with legal cases relating to internal relations within a company or organization.

On April 9, 2026, the Financieel Dagblad reported that, to the indignation of the judges, the lawyer for a minority shareholder had filed a petition containing incorrect ECLI numbers and, contrary to the truth, described the minority shareholder's role as smaller than was factually the case in this dispute.

AI literacy among lawyers

Since February 2, 2025, it is a statutory obligation to promote 'AI literacy' in your company or organization. Law firms included, these days law firms also must ensure an adequate level of AI literacy among their staff and other persons who operate and use AI systems on their behalf as deployers responsible for AI systems.

We previously wrote about the approach to getting a grip on AI within your organization:

deel ii ai recht | hoe maak ik ai beleid voor mijn eigen bedrijf of organisatie 

Earlier, the Dean of the Gelderland Bar Association intervened by sending two lawyers who had improperly used generative AI—by submitting procedural documents containing non-existent court rulings—to a 'mandatory AI course'. Only time will tell whether it will remain limited to two lawyers.

A donkey rarely stumbles...

Making mistakes is human and inevitable, especially when using completely new technology as a tool for the first time. Proponents of embracing innovation despite the risks will argue that making mistakes is simply part of the learning process.

However, a lawyer in Rotterdam raised eyebrows when he was reprimanded by the same court for a second time after submitting a legal document full of fake case law hallucinated by AI.

District Court of Gelderland 6 November 2025, ECLI:NL:RBGEL:2025:9423, paragraph 3 (last sentence)

"The court cannot escape the impression that use was made of ChatGPT or another AI tool. Be that as it may, given that some (alleged) rulings do not exist and the relevance of other cited rulings is unclear, the court will disregard the reliance on this (alleged) case law."

District Court of Gelderland, 4 March 2026, Rb ECLI:NL:RBGEL:2026:1608, ro 3.1

"The court cannot escape the impression that the plaintiff’s representative (again) made use of ChatGPT or another AI tool when drafting his grounds for appeal. Although the plaintiff’s representative was alerted to this in the court’s earlier ruling, he did not choose to withdraw his letter containing grounds for appeal and replace them with grounds for appeal referring to relevant case law, or to have the irrelevant rulings removed from his grounds for appeal. Be that as it may, given that the relevance of some of the cited rulings is unclear, the court will disregard the reliance on this (alleged) case law."

Future predictions: consequences of the correct use of AI in legal proceedings?

Whether due to the growing attention to checks to prevent fake ECLI numbers or not, there is a lawyer who was able to indicate that the sources in a ChatGPT- generated response had been verified. However, the AI-generated piece of evidence ultimately did not hold up after further questioning by the judge.

District Court of East Brabant, 14 April 2026, ECLI:NL:RBOBR:2026:2232, ro 4.6 – 4.7

"The explanation included by the lawyer on this point in the statement of defense is based on what ChatGPT has written and supplemented by the arguments of UMS. The lawyer has indicated that his client has verified the sources cited by ChatGPT in the analysis."

In this case, the lawyer was unable to provide sufficient answers to questions regarding the manner in which ChatGPT was used; in particular, he could not demonstrate which prompt had been used, could not indicate which ‘temperature setting’ he had used, and left the impression that he did not understand the question, which could have been interpreted as a lack of technical knowledge or ‘AI literacy’ that disqualified the persuasiveness of the AI-generated piece of evidence. Moreover, it emerged that a version entered during the input was an older draft from an earlier date and not the final version with attachments that was before the court in the case. Taking all circumstances together, the court decided to disregard the document, but only after the opportunity to convince the judge of its reliability had been given:

"This has consequences for ChatGPT's response. Given this state of affairs, the court will disregard exhibit 3, and the arguments the lawyer has put forward regarding this."

ChatGPT has not yet won any court cases in the Netherlands, although that day may be drawing closer.

We keep a close eye on developments surrounding AI law for you.

Questions?

Are you seeking legal advice regarding what the rise of AI means for you or your organization? Do you want to know more about the legal aspects of AI? Have you drafted an AI policy and would you like to request a legal review?

Contact our IT legal experts and lawyers today. They are happy to assist you!