What Happened
The EU is betting on AI to revolutionize healthcare with initiatives like the €21B Horizon Europe fund and Germany’s €500M AI Innovation Competition. Yet actual deployment lags. DeepMind’s AlphaFold 2 solved protein folding but EU startups like Owkin or BenevolentAI still struggle to turn models into approved therapies. Brussels talks of a ‘digital decade’ but its AI Act’s compliance costs could stifle the very startups it claims to nurture.
Why It Matters
AI in drug discovery is a tool not a miracle. The EU’s scattergun approach funds pilot projects but lacks the capital intensity of US peers. Flagship programs like the European Health Data Space aim to unify records by 2025 yet member states resist data sharing. The real bottleneck isn’t algorithms but access to patient data and late stage trial funding where the US outspends Europe 3 to 1.
Who Wins & Loses
US Big Pharma and Big Tech win by acquiring EU startups on the cheap. European patients lose as breakthroughs get priced out or exported. Owkin and BenevolentAI risk becoming features not companies. Siemens Healthineers and Philips may leverage AI for diagnostics but lack the scale to dominate therapeutics.
What to Watch
Watch for EU attempts to mandate data sharing under the Health Data Space which will face lobbyist pushback from hospitals and insurers. Expect more EU startups to pivot from cure all claims to niche diagnostics where regulatory hurdles are lower. The first AI designed drug to reach market in Europe will be a litmus test likely delayed until 2026 or later.
Social PulseRedditHackerNews
European engineers are skeptical but politically quiet. The hype cycle has them chasing grants not breakthroughs. Founders privately admit the best exit is acquisition by Pfizer or Novartis. The chatter reveals a brain drain risk as talent follows capital across the Atlantic.
Sources
- Can AI really cure all diseases?