NSS opakovaně varuje před nekritickým využíváním AI při přípravě podání

In a recent judgment, the Supreme Administrative Court (SAC) highlighted the risks associated with the use of artificial intelligence in preparing cassation complaints. Although the Court was not deciding on the use of AI itself, it devoted a significant part of its reasoning to the quality of the cassation complaint, which, in its view, exhibited characteristics of AI-generated text.

The SAC noted that the filing contained numerous inaccurate or entirely non-existent case law citations. Some decisions were cited out of context, others contained incorrect dates, and one of the cited decisions did not exist at all. The Court described part of the argumentation as likely generated by AI and even referred to it as “AI slop”.

The judgment follows previous decisions of Czech courts that have been critical of the uncontrolled use of AI in legal practice. The SAC reiterated that responsibility for the content of a filing always lies with the person who signs it. Professionals providing legal services therefore bear full responsibility for any AI “hallucinations” and the procedural errors that may result from them.

This latest judgment serves as a further reminder that AI can be a useful tool for processing information and preparing supporting materials. However, particularly in the field of taxation, the highest level of accuracy remains essential. Even a seemingly minor error or an unverified AI-generated output may lead to significant procedural, financial, or reputational consequences.

Authors:

Radek Matuštík
Jiří Dienstbier