On a swing through Europe, Jensen Huang committed $2B in new investments across the continent, per Bloomberg, and used a Paris stop to reveal fresh stakes in the French startups Doctolib, Genesis AI, Pigment and Entalpic. Because Nvidia already spent about $8B in Europe last year, the move continues an established pattern rather than starting a new one.
The Nvidia chief spent the trip arguing that open-weight models are Europe’s quickest path to narrowing its AI gap — a pitch pointed straight at Brussels as it weighs sovereignty rules for cloud and AI providers.
Compute for the safety crowd
British researchers got a pledge of their own. A five-year package of compute worth $250M, aimed at AI agent safety and security work, is going from Nvidia to named recipients: Oxford’s BOLD group, UCL’s SOFAIR project and Imperial College London. Topics in scope for the three include robotics safety, open-source architectures and open-ended learning.
UK AI Minister Kanishka Narayan described the commitment as a vote of confidence in the country’s research base.
Buying influence as rules tighten
Nothing about the timing is accidental. The Cloud and AI Development Act is under consideration in Europe, and it could channel public procurement, then private workloads, toward providers with European ties. Whoever operates them, every major cloud rests on Nvidia hardware — but stakes in local startups and free compute for safety researchers purchase something harder to legislate: goodwill among the officials drafting the rules and the researchers who critique them.
