Can Europe trust Google with its AI? I sat down with Google VP of Engineering Wieland Holfelder to discuss sovereign AI, cloud infrastructure, and who stays in control when AI agents start taking action. YouTube in comments. Have fun :)

@Google Chubby on YouTube! Let's goooo!!

@Google You guys are so nice :))

@Google oh hell yaaa very nice

@Google @kimmonismus what's tricky is ensuring data privacy while scaling AI. It's a massive trust factor in Europe.

@Google This is such a useful angle. The control question feels especially real as agents start acting on their own.

@Google The agent part is where “control” gets concrete for me: being able to say “finish this task, but these actions are off limits.”

@Google Didn’t know you create videos! I’m sat!

@Google Just started :))

@Google sovereignty question gets concrete the first time an agent makes a wrong call and you're piecing together why from three vendor dashboards. turns out 'control' is just whether you own the complete action log

@Google YES! We’re getting Chubby videos for real 😁👏🫶

@Google ❤️❤️

@Google I'd be curious how they define sovereign AI beyond data residency.

@Google trust is a function of where inference runs and who holds the keys, not where the vendor is registered. europe can check that part

@Google Yes. Their European cloud regions keep the data on local soil, but Google’s algorithms still rule the model layer. It’s a comfortable dependency, like letting them handle the laundry while you decide what goes in which drawer. Efficient? Yes. Sovereign?

@Google @grok что говорится в этом видео? Я просто не знаю Английский и не могу понять что там говорится

@Google my AI asks before adding people to a group chat. the bar is apparently underground

@Google control over ai agents is the real question rn, big topic