S&D group calls for EU-wide ban on harmful non-consensual deepfakes

The Socialists and Democrats group in the European Parliament is pushing for EU-wide legislation to ban harmful, non-consensual and deceptive deepfakes, requiring platforms and AI providers to prevent their spread rather than respond only after victims file complaints.

A draft of the group's tech policy objectives, seen by journalists in Brussels, sets out the proposal as part of a broader Europe-wide debate over control of digital identity as AI enables easy production of fake political endorsements and voice clones used in scams.

"The current trend towards different national solutions only creates legal confusion bad actors could exploit," said German S&D MEP Tiemo Wölken, a key legislator on the Parliament's copyright and AI report.

The push follows a report that 50 MEPs have been targeted by sexualised deepfakes. So-called nudification apps are due to be covered by an update of the AI Act, but the S&D proposal targets the broader deepfake problem. The draft does not specify the mechanism for implementing the ban.

Denmark this month introduced a plan to amend its copyright law to bar sharing digital replicas without consent, targeting public distribution while allowing exceptions for parody and satire. Italian Prime Minister Giorgia Meloni separately applied to trademark a recording of her voice against unauthorised use.

Some legal experts warn that building new protections into copyright could backfire. Stina Teilmann-Lock, an associate professor at Copenhagen Business School, cautioned against this approach, saying that "baking it into copyright" would create problems for copyright law and might not solve the underlying issue.

Existing EU rules — the AI Act's transparency requirements and the Digital Services Act's illegal-content provisions — do not create a general EU-wide prohibition on deepfakes.