Google is handing Europe's scientists the keys to its lab. On Wednesday, Google DeepMind and Google Cloud announced they will give up to 200 European researchers direct access to frontier AI models, alongside $3 million in compute credits, to accelerate scientific research across the continent.
The researchers will be chosen by the European Commission under its Resource for AI Science in Europe (RAISE) scheme. The toolkit on offer is serious: AlphaFold 3 for protein research, WeatherNext for weather forecasting, plus AlphaGenome, AlphaEvolve, and AlphaEarth Foundations. Google framed the move as a way to "accelerate the pace of European innovation and reinforce the continent's scientific competitiveness," per the statement released at an AI summit in Dublin.
The timing is deliberate. Europe has been pressing for stricter AI regulation — demands that have only grown after a string of security incidents — while simultaneously worrying that its researchers are falling behind the US labs on compute access. Opening frontier models directly to handpicked academics is Google's answer to both pressures: goodwill plus a foothold in the region's scientific ecosystem.
This follows a pattern. The labs that once kept frontier models behind closed APIs are increasingly courting public science — researchers get compute and model access, the labs get credibility and scientific validation of their tools. For European labs tackling everything from extreme-weather prediction to molecular structure mapping, the compute credits may matter as much as the models themselves.
As always with such announcements, the fine print — who exactly gets chosen, under what terms, and how much real compute $3 million buys against today's workloads — will determine the impact. But for European science, today was a good day.