Italy’s competition authority, the AGCM, has opened an investigation into Suno, the US company whose AI service turns text prompts into songs.The authority said today that parts of Suno’s terms of service may be unfair to consumers under Italian law.

The case is based on Article 33 of Italy’s Consumer Code, and according to this article, a contract term is considered unfair when it tips the balance of rights too heavily in favour of the company and against its customers.

The AGCM raised a number of concerns, pointing out that Suno’s terms seem to allow it to alter the contract, the service, and the subscription prices on its own without providing a reason.

Furthermore, the terms seem to permit the company to suspend or close accounts, along with any songs stored in them, at any time and without notice.

Moreover, when users are signing up, they are apparently asked to agree to further terms which they cannot see at that time, and the terms seem to restrict Suno’s liability in general terms, the authority said.

Copyright is the point of special attention. To use the platform, users must grant Suno a licence to use their content, and the AGCM stated that this licence is defined too broadly.

Furthermore, the terms require users to waive their moral rights, which allow authors to control how their work is credited and treated. Moral rights are protected by Italian copyright law.

The authority also raised questions about the way disputes are dealt with, since Suno’s terms seem to have consumer claims referred to binding individual arbitration in the United States, with strict time limits and no provision for class actions; they also state that the courts in Massachusetts, the state where Suno is based, have exclusive jurisdiction.

In the weeks to come, the AGCM will hold a public consultation on the contested terms through its website, and national trade associations, chambers of commerce and recognised consumer groups will be allowed to take part.

This case represents the latest instance of the AGCM taking consumer action against an AI company.

Earlier this year, the organisation concluded its investigations into DeepSeek, Mistral and Nova AI after the three companies agreed to alert their users to hallucinations, which occur when chatbots generate false information.