Op-ed on 'Europe’s Advanced AI Strategy Depends on a Scientific Panel. Who Will Make the Cut?'

The European Commission is setting up one of the most consequential governance bodies under the AI Act: the AI Scientific Panel. Its 60 independent experts will serve a two-year term, advising the European AI Office (and national authorities) on GPAI models and systems. They will have an big say on how the most advanced AI technologies need to be tested or what future developments would mean for the law.

This panel will be more than a “nice-to-have.” It is meant to help Europe to get the hard things right: assessing systemic risks, classifying models, shaping evaluation methodologies, and supporting cross-border market surveillance, while also keeping an eye on emerging risks that may not yet be visible today.

But there is a catch: Member States have significantly constrained the Commission’s room for maneuver by pushing for national quota. Each of them must be represented, and 80% of seats are reserved for experts from the EU or EFTA. That may sound politically reassuring, but it raises a simple question: will this approach reliably deliver the best expertise for the world’s most advanced AI systems or will it dilute excellence at the very moment Europe needs it most?

My view is straightforward: the selection should be driven by merit and proven frontier expertise, not passports. Europe should aim for world-renowned researchers and genuinely independent third-party evaluators; people who test, stress, and scrutinize advanced models beyond corporate talking points. And it should bring in younger voices with the capacity and time to do the work, not just senior profiles with impressive titles.

If Europe wants an “advanced AI strategy,” it has to start with an advanced choice: prioritizing scientific excellence, independence, and real-world relevance over geography, optics, and box-ticking.

Read the full piece on TechPolicy.Press here.