AI and Image Rights: The Debate on Commercial Deepfakes

1. The facts

   According to press reports, a well-known Italian TV host publicly complained on her social media channels about the unauthorized use of her image by a wellness company. The case did not result in any emergency petition or judicial act, but concerned the distribution on social networks of retouched photos and promotional videos in which the protagonist appeared to recommend weight-loss supplements. The case sparked a wide debate on what would happen if, instead of simple retouched photos, a full deepfake created via generative AI were used, capable of perfectly replicating physical features, lip movements, and vocal timbre, making the person utter unauthorized statements.

2. 

   The applicable framework

   The central provisions for the theoretical resolution of the case are three: first, Article 10 of the Italian Civil Code protects against the abuse of another person's image, allowing the court to enjoin any display or publication of a portrait made without consent or to the detriment of dignity; second, Articles 96 and 97 of Law no. 633/1941 establish that a person's portrait cannot be displayed without authorization, except in specific cases of public notoriety which must never harm the person's honor; third, Articles 4 and 9 of the GDPR classify the face and voice as biometric data, a special category of sensitive personal data whose processing requires strict legal bases and explicit consent, especially when aimed at unique identification or commercial exploitation.

3. What case law says

   Supreme Court case law has long clarified that the right to one's image is an absolute right and that its commercial exploitation almost always excludes the applicability of exemptions related to news reporting or public notoriety. Established case law equates the protection of voice with that of the image, considering it a fundamental distinctive trait of personal identity. Lower courts have further specified that the infringement exists even if the image is artificially created, as the object of protection is not the original photographic material, but the social projection and identity of the person improperly evoked and manipulated.

5. Analysis drafted and verified with edit.legal

   To verify the provisions cited in this article, we used edit.legal. Test our legal AI on official sources and apply it to your own matters.

   Try edit.legal AI

8. What it teaches professionals

   1. Always verify the chain of consent not only for archive material, but also for datasets used in training generative AI models for commercial purposes.
   2. Include specific clauses in image licensing contracts that expressly exclude or regulate the creation of synthetic avatars or vocal clones.
   3. Act promptly in summary proceedings as soon as a deepfake is detected, as the virality of digital content accelerates image erosion damage.
   4. Consider the protection of biometric data under the GDPR as an additional procedural tool, often more effective than claiming copyright infringement alone.

9. Update and rectification note (17 September 2026)

   The previous version of this article incorrectly reported the existence of a summary injunction by the Court of Rome regarding a deepfake of a well-known TV presenter. Following an editorial review, we confirm that there is no judicial act nor any legal precedent established in this regard: the real case is limited to a social media complaint about the use of retouched photographs. The text has been corrected to reflect the absence of legal proceedings and to frame the issue as a theoretical debate.

References: Art. 10 c.c.Art. 96 L. 633/1941Art. 97 L. 633/1941Art. 4 GDPRArt. 9 GDPRArt. 700 c.p.c.

Editorial oversight: Avv. Federico Papa·ICAMUpdated 17 September 2026Content drafted with AI support and subject to editorial source checks. Despite these controls, inaccuracies may remain: reports and rectification requests are welcome. Report a correction