The Irish Council for Civil Liberties (ICCL) has lodged a formal complaint with the European Ombudsman, alleging that the European Commission has used generative artificial intelligence (AI) in public documents without sufficient transparency.
The complaint follows the ICCL’s review of a recent response from the commission to an access to documents request. The council identified that at least one link in the documents appeared to have been generated using OpenAI’s ChatGPT.
ICCL officials warned that generative AI tools can sometimes produce inaccurate information, raising concerns about the reliability of official EU documents. “Public bodies like the European Commission should always be transparent and disclose if a generative AI tool is used in any public document, even if the output from such tools has been assessed by their staff,” said Dr Kris Shrishak, ICCL Enforce Senior Fellow. He added that disclosures should include detailed information about the AI tools used to ensure accountability.
The ICCL’s action highlights growing scrutiny over the use of AI in government and institutional communications across Europe. Critics argue that without clear disclosure, citizens may not be aware when automated systems contribute to official messaging or documentation, potentially undermining trust in public institutions.
A spokesperson for the European Commission responded, saying the institution has clear internal guidelines for AI use. “We have our own GPT tool,” the spokesperson said. “We’re using AI in our daily work, but there is always human oversight.” The commission emphasized that its AI practices are guided by responsible use protocols and that its internal guidelines were most recently updated in May.
The ICCL complaint comes amid broader discussions at the EU level about the regulation of generative AI and the ethical responsibilities of public institutions in deploying these technologies. Transparency and accountability have been central themes in debates on AI, particularly as these systems become more integrated into administrative and decision-making processes.
Legal and digital rights experts say the outcome of the complaint could set an important precedent for how EU institutions disclose AI use in public-facing documents. It may also influence future policies on the verification and monitoring of AI-generated content in official communications.
Dr Shrishak urged the commission to adopt stronger disclosure practices to maintain public confidence. “People have a right to know when AI has been used, especially in contexts where accuracy and accountability are critical,” he said.
The European Ombudsman has yet to comment on the ICCL complaint, and it remains unclear what steps the office may take to investigate the use of AI in EU documentation.
