Scrutinising Strategic AI Partnerships: Win-Win or Winner-Takes-All?[*]This article has been peer-reviewed.

By Ruben Verdoodt[**]Ph.D. Researcher at KU Leuven, Consumer Competition Market.

1. Introduction

The booming of the artificial intelligence (AI) industry has gone hand in hand with the rise of a peculiar commercial phenomenon. AI companies regularly enter into partnerships with vertically-integrated technology firms such as Alphabet, Amazon, Meta, Microsoft and recently also NVIDIA (collectively referred to as ‘major technology firms’). They do so to access crucial inputs necessary for AI development, such as computing power, capital and vast amounts of data. In return, technology firms can demand product licenses, customer commitments, information and decision rights, as well as skilled labour through so-called acquihires. Despite an apparent procompetitive win-win character, the power imbalance underlying the partnerships also gives rise to legitimate competition concerns.

The article pursues several objectives. In the first place, it provides a comprehensive description of the phenomenon of strategic AI partnerships, their characteristics and the threats to competition which they may pose. In this regard, the article discusses traditional theories of harm as well as wider market dynamics and contestability considerations. The issue of acquihires is also studied in detail. In a second step, the article then turns to the legal arsenal at the disposal of the competition authorities in Europe (EU, Member States, and the UK) to address AI partnerships. The aim here is to provide an inventory of enforcement options, while measuring their theory against practice. Throughout the article, reference is made to real-world partnerships, including those under scrutiny.[1]Updated until 1.3.2026. In the conclusions, two avenues are considered for guiding the way forward.