The story mixes dizzying numbers with uncomfortable questions. Neko Health, the body-scanning startup founded by Spotify’s Daniel Ek, has just closed a $700 million funding round, putting its valuation at nearly $7 billion. The company is preparing to land in New York, backed by investors like Mark Zuckerberg, Maria Sharapova, and will.i.am. The pitch: offer comprehensive preventive body scans for a fee, even in markets where healthcare is free. As the company put it, “We’ve gone into markets where health care is free, and hundreds of thousands of people line up in a queue” to pay for the service.

But beyond the excitement for prevention that’s accessible to those who can afford it lies a knot tied to data sovereignty. Each scan produces a staggering amount of sensitive information: diagnostic images, physiological markers, likely genetic data. Where does that data end up? Who processes it and with what guarantees? In Europe, GDPR imposes strict rules on health data processing, and any cross-border transfer – especially to the United States – demands adequate safeguards. The New York expansion, just as U.S. privacy regulation could tighten, only amplifies the dilemma.

Managing clinical data flows at scale isn’t an infrastructural afterthought; it’s the core of Neko’s business model sustainability. If scans are processed in the public cloud, the risk of breaches and unauthorized access grows in direct proportion to the user base. That’s why more and more healthcare operators are evaluating on-premise or edge architectures, where data remains physically under the control of the service provider – or better yet, the patient. Such a framework would reduce long-term TCO, avoid cloud egress costs, and, crucially, strictly meet data residency requirements.

Within the AI-RADAR ecosystem, which precisely analyzes trade-offs between cloud and local infrastructure for workloads, the Neko case is emblematic. This isn’t a hypothetical LLM fine-tuning scenario; it’s an industrial reality that demands high-performance servers – from the compute power for image analysis to the VRAM needed for diagnostic AI models. Opting for an entirely cloud-based deployment could set the company on a collision course with data protection authorities, rapidly eroding user trust.

It remains to be seen whether Neko Health can reconcile its aggressive commercial expansion with data governance that meets regulatory expectations. Zuckerberg’s presence, a long-time advocate of cloud-centric approaches and interconnected platforms, offers little comfort to those who fear a drift toward data concentration in a few hands. On the contrary, it may trigger an even fiercer debate about the wisdom of keeping health data on-premise, just as the company sets out to conquer New York.