On January 5, 2026, the Italian Competition Authority (Autorità Garante della Concorrenza e del Mercato – AGCM) closed the investigation initiated against the Chinese artificial intelligence company DeepSeek, without establishing any infringement. The decision followed the acceptance of commitments proposed by the company, which were deemed sufficient to eliminate potential unfairness concerns related to the commercial practices under investigation, as reported in the authority’s official bulletin.
The proceeding focused on the alleged lack of adequate information provided to users regarding the possibility of so-called “hallucinations” in AI models, namely situations in which the system generates inaccurate, misleading, or fabricated outputs in response to user inputs. The commitments undertaken by Hangzhou DeepSeek Artificial Intelligence and Beijing DeepSeek Artificial Intelligence, which jointly own and operate DeepSeek, consist of a set of measures aimed at improving transparency and user awareness about the risks associated with such AI-generated outputs.