The President of the European Parliament, Roberta Metsola, has formally asked the European Commission whether it uses AI in its responses to MEPs, according to Politico.
This comes after a complaint by the President of the “Renew Europe” group, Valérie Hayer, who said that the European Commission’s responses to MEPs are “of poor quality” and “delayed.” The European Commission is legally obligated to respond to such inquiries, but is not bound by specific timeframes or quality standards.
Some members of Parliament argue that the European Commission often ignores or overlooks the European Parliament, creating frequent tensions, while others accuse Commission President Ursula von der Leyen of “disliking the European Parliament.”
At the initiative of the European Commission, its staff use a separate version of ChatGPT (GPT@EC) for internal use only (e.g., updates).
Artificial intelligence “is currently being tested, for example, for correctly formulating questions between the Commission’s services or for identifying previous responses that might be relevant to a specific case,” said a European Commission spokesperson.
According to the spokesperson, staff are not allowed to “reproduce the output of an AI model in public documents, such as creating Commission texts, especially those that are legally binding.”
However, it is already being used for drafting, summarizing documents, brainstorming, or generating software code.