Does one company's massive fundraising round truly inform the general viability of the AI Act? For most startups, Mistral AI's success is a single point of failure as it doesn't reflect the daily grind of compliance costs.
Many smaller Belgian AI companies will still find the legal framework a weakest link, regardless of how well an outlier performs.
For example, my neighbour's son runs an AI startup in Ghent; they're spending more on lawyer fees for data governance this year than on new server hardware.
That kind of blast radius isn't covered by one big success.