European Commission AI Guidelines Breach

The Irish Council for Civil Liberties (ICCL) has filed a formal complaint with the European Ombudsman regarding the European Commission’s recent use of generative AI, specifically OpenAI’s ChatGPT, in public documents. This action potentially breaches the Commission’s own established guidelines and its obligations under EU treaties, raising significant concerns around transparency and accuracy.

Recent correspondence from the European Commission, responding to an ICCL request for access to certain documents, unexpectedly revealed that AI-generated content was included in their public communications. Among the documents, at least one link pointed directly to material produced with ChatGPT, as indicated by the presence of “utm_source=chatgpt.com,” highlighting a concerning reliance on generative models.

The implications of utilizing AI systems for crafting public documents are noteworthy. According to the Commission’s own staff guidelines, there is a strict prohibition against directly replicating outputs from generative AI models in official communications. The nature of AI is such that it produces content based on probabilistic algorithms, which can lead to the dissemination of information that is not only inaccurate but also misleading—a significant concern for institutions tasked with providing reliable information to citizens.

Dr. Kris Shrishak, the ICCL Enforce Senior Fellow, emphasized the importance of transparency when public bodies opt to employ AI tools, asserting that, “Public bodies should always disclose the use of generative AI tools in any public documents.” He contends that such transparency is crucial, given the unpredictable nature of AI-generated content. Furthermore, if used, the responsibility should lie with the institution to confirm the correctness of the information rather than placing the onus on the public to challenge misleading claims.

This situation raises broader questions about the role of AI in governmental communication and the need for stricter compliance with ethical standards in AI’s application within public institutions. As discussions around digital rights and AI governance evolve, the actions of the European Commission may serve as a critical case study in maintaining accountability and preserving the integrity of public administration.