Skip to content

Balancing Competing Interests in the Reidentification of AI-Generated Data

Emmanuel Salami


AI systems generate valuable analytical information from (apparently) non-personal data with vast economic consequences. This information generated from non-personal data provides a competitive edge which serves as a key rationale for its appropriation to the exclusion of others. The proliferation of AI has made it possible for non-personal data (including anonymised data) to result in the reidentification of natural persons. There have been propositions from various quarters for the protection of non-personal data by Intellectual Property Rights (IPR). Should AI-generated data be protected by IPR, this can potentially result in data protection concerns in the event of data reidentification that singles out natural persons. This might particularly occur where reidentified data leads to the identification of natural persons in circumstances where the applicability of data protection law had neither been contemplated nor anticipated. This article highlights the concerns that might arise in the event of data reidentification and how this might raise interesting challenges for data protection compliance.
Keywords: AI | AI-Generated Data | Reidentification | Intellectual Property Rights

The author is a PhD candidate and researcher at the University of Lapland, Rovaniemi, Finland. He is also an in-house data protection advisor for a global tech company. Any views expressed in this article are those of the author only. For correspondence: <>


Lx-Number Search

(e.g. A | 000123 | 01)

Export Citation