Abstract

Discussions of digital sovereignty predominate artificial intelligence (AI) discourses. However, digital sovereignty has been unable to effectively respond to longstanding concerns regarding the use of AI. These challenges include systemic bias, transparency/accountability and intellectual property infringement/theft. The authors posit an alternative framework – informational sovereignty – encouraging a recalibration of how technological sovereignty is viewed. Through this model emphasis is placed on respecting jurisdictional boundaries and jurisdictionally appropriate information sources to result in representative outcomes for communities rather than the traditional focus on where the data is held and system reliability that has thus far been the subject of much high-profile litigation. The article therefore sets out a quadripartite model of informational sovereignty encompassing concerns regarding population, territory, recognition and regulation of borders, before analysing the place of informational sovereignty in future iterations of AI regulation, including its practical applicability in the European Union Artificial Intelligence Act (EU AI Act).