6 Comments
User's avatar
Katherine Lazarevich's avatar

Thank you for the post! I care about Europe's future when it comes to AI. I live here and run an AI business here. On one hand, I comfort myself that Europe often adopts slowly, but does it thoroughly. In Poland, contactless Blik payments are accepted everywhere, while three years ago on Broadway in New York they wouldn't take my Apple Pay 🙂

On the other hand, regulation doesn't exactly thrill me as a business owner. One of our products is AI voice, and we're working on rollouts with several insurance companies right now. Insurance is a regulated industry with everything that follows, and it's all moving painfully slowly.

But what worries me even more is that we don't have a frontier model of our own. We're not part of the US–China race at all. And there have already been cases where the US administration didn't want to share models with anyone outside the States. I'd very much like us not to end up on the sidelines.

Silvia Stepitova's avatar

Thank you, Katherine. I love the Blik example. Slow and thorough is the optimistic version, and sometimes it's true. On insurance, I'd say it's a combination. Insurers are building their AI governance right now, so every new AI tool lands in the middle of that, on top of the usual vendor reviews: data location, subprocessors, security, outsourcing. Arriving with answers to both helps. That means which risk category your product falls into, how it tells callers they're speaking to an AI, and where the data goes. On frontier models, I share the worry.

Katherine Lazarevich's avatar

Yes, we're in the middle of this "journey" with insurance 🙂 And your list is spot on: where the data goes and how callers know they're talking to an AI come up first. I hope for an optimistic future, but sometimes it feels like Europe is a bit asleep.

Weekly Sovereign AI News's avatar

Thank you for sharing the Amazon story so openly; it gives the numbers real weight. The EIB figures (37% vs 36%) suggest adoption isn't the gap, it's what happens after the pilot. From what I see, the slowest approvals in European companies are less about the AI Act itself and more about unclear answers on where data goes and who can access it. Did the Slovak round table read the parent-company approval loops as legal caution, or as uncertainty about the vendors?

Silvia Stepitova's avatar

Thank you. The round table didn't go that far. People described the waiting more than its cause. I would say it’s a bit of both. AI Act is new, so companies are cautious, and they're also still working out what exactly they need from vendors.

Mean Cos's avatar

So you did have an answer