Mapping Sovereignty and Competition Concerns Along the Generative AI Value Chain

Competitive, technological and data advantages in the platformeconomy have offered leading digital firms, such as Google, Amazon, Meta, Microsoft and Apple (GAMMA), a competitive lead in the generative AI race for supremacy. Recent innovations by the large digital players, such as the ‘word vectors’ and ‘transformers’ by Google and innovative developments in graphics processing units by Nvidia, have contributed to rapid developments in the industry. Emerging champions in generative AI are either the incumbent digital gatekeepers or, if a new entrant (OpenAI), have benefited richly from the support provided by the current gatekeepers. Discernibly, a handful of digital oligarchs repeatedly influence each successive wave of innovation and design standards for the critical technological infrastructure. This contribution decodes the complex value chain of generative AI and maps how these competition concerns also qualify as digital sovereignty along the digital stack. The discussion looks at the AI infrastructural layer and self-preferencing in search. Appreciation of the underlying policy rationale of a regulatory measure can facilitate more effective implementation. The discussion maps generative AI on the competition law and Digital Market Act’s regulatory landscape, and illustrates the latter’s hitherto implicit sovereignty goal, alongside its more widely accepted pursuit of fairness and contestability.