# Why Europe’s demands on Apple AI put your data at risk > Source: > Published: 2026-06-15 15:21:51+00:00 Europe’s [evangelistic approach](https://www.applemust.com/apples-siri-ai-stand-off-with-europe-just-escalated/) to insisting Apple open up personal data to competing AI services is hurting Apple users in the region. More than that, it also places its entire business sector at risk, and a [newly-published Jamf survey](https://www.businesswire.com/news/home/20260615806745/en/Jamf-Survey-finds-AI-incident-rates-rise-as-organizations-deepen-AI-integration) suggests why. [Announced at WWDC 2026](https://www.computerworld.com/article/4182977/wwdc-what-analysts-and-the-tech-industry-say.html), Apple Intelligence/Siri AI relies on personal, contextual data to run. Europe wants that same information to be made available to third-party services for competing apps, but has not worked with Apple to protect user confidentiality. It’s an approach that places your data at risk of exfiltration using those apps because Europe is insisting Apple share personal information with the developers of other apps. The desire to protect that data is why Apple won’t distribute Siri AI in the EU for a while. It’s not as if Europe doesn’t understand the risk of data leaks in an era of AI. Just look at the bloc’s focus on things that *do* matter, such as sovereign AI or managed AI services like Orange Live Intelligence. These locally-produced AI services, alongside Europe’s attitude toward them, tell me the confederation understands the risks. How real are these risks? Very. Jamf on Monday published survey results confirming the scale of that risk, telling us that one-in-five IT and security leaders in the enterprise sector has already experienced an AI-related incident involving unexpected costs, a security issue, or both. The survey also found that: The implication is that AI governance is becoming an operational requirement and — as Apple has told us umpteen times in the past — the best way to maintain operational confidentiality is not to collect or share any data at all. That’s the whole point of its approach: the data doesn’t need to be shared, it just needs to be turned into another signal that promotes utility while protecting confidentiality. There’s another challenge to emerge. There are now multiple brands of AI, with more coming on stream all the time. That’s great in terms of finding a model that suits your needs, but challenging when it comes to ensuring all the services you or your employees use of are equally secure. You don’t want your business to become deeply reliant on any service only for that vendor to subsequently get bought out and/or shut down, nor do you want a service to be hacked or otherwise exploited to your detriment. “AI isn’t arriving as a single application that IT can approve and move on from,” said Jamf CEO Beth Tschida. “It’s showing up in developer tools, productivity apps, autonomous agents, and other software they already run. The challenge is maintaining visibility and control as that footprint expands.” The survey described the challenges IT faces with AI deployment: shadow IT, vendor sprawl, and the need to grapple with highly unpredictable use-based pricing models. And that’s even before considering the governance challenges of agentic and developer AI. “What our survey shows is that governance must keep pace with adoption,” Tschida said. “For organizations built on Apple, the foundation is already an advantage. Apple’s privacy model and the management controls built into the platform give IT teams a strong foundation to build on and … that advantage depends on using tools built for Apple from the start.” That’s the point of the curated, private and secured service offered by Siri AI, of course. It’s also part of what Apple is building toward with its wider ambitions toward AI on its platform. * Bloomberg’s* Mark Gurman discussed elements of this in his weekend newsletter, in which he suggested Apple might introduce some subscription services using AI, and that it is building an App Store for Siri Extensions, which would What makes that model work is the curation with which Apple surrounds it, and its determination to extend Private Cloud Compute so it can protect your data even when using third-party servers (in this case, Google’s server clusters). It makes sense to think Apple intends to use that system to protect all approved third-parry AI interactions provided across its platforms by its own routes. That’s true, even if users access services free of those safeguards using a web browser, which they currently can. But the key thing is that if Apple can get this right, offering up a managed, curated, and controllable ecosystem for AI agents and services, it will be going a long way toward building the kind of managed AI ecosystem the Jamf survey shows our modern digital enterprises increasingly need. *You can follow me on social media! Join me on BlueSky, LinkedIn, Mastodon and subscribe to The Core.*