Governments around the world are ramping up efforts to regulate and govern AI systems. Trialogue negotiations on the AI Act are currently underway in the EU, with the expectation that EU policymakers will agree on the final text of the Act in the coming weeks. On the other side of the Atlantic, on October 30, US President Joe Biden issued the AI executive order, which addresses the wide range of issues related to AI benefits and risks.
Against this backdrop, on November 1-2, the United Kingdom hosted a high-profile AI Safety Summit. The meeting brought together representatives from governments and big tech (with a notable exclusion of communities and workers most affected by AI).
At the meeting, governments from the United States and the United Kingdom pledged to invest in the computing infrastructure required to develop AI that serves the public good. It is concerning and puzzling that the European Union has not committed to this crucial effort. While the United States and the United Kingdom are taking a leading role, the European Union is falling behind.
Ursula von der Leyen recently expressed her support for digital public infrastructures that are trusted, interoperable, and accessible to all. In this context, investment in computing power that would allow independent researchers to study and develop AI is a must for ensuring Europe’s digital sovereignty. The EU has been the trailblazer for curbing the dominance of big tech through legislation. While regulation is important, the democratization of AI development hinges on public investment in the AI infrastructure.
Experts agree that artificial intelligence is the next frontier of market concentration in the Internet economy. We’ve previously pointed out that one of the main bottlenecks to AI democratization is dependency on private infrastructure. To reduce market concentration and AI development dependency on big tech, states must invest in public infrastructures. In the context of AI, this means establishing a robust and publicly accessible computational foundation for open-source AI research.
