Apple is enhancing transparency around its Apple Intelligence features by adding a new user permission popup that notifies and seeks consent before certain AI tasks send data to Google Cloud servers. The change appears in both the upcoming iOS 27 and the current iOS 26, 9to5Mac reports, reflecting Apple’s expanding use of external cloud infrastructure for advanced generative AI capabilities.
Background on Private Cloud Compute Evolution
When Apple Intelligence debuted in 2024, the company emphasized Private Cloud Compute (PCC) as a privacy-focused innovation. PCC originally ran exclusively on Apple’s own secure servers, earning praise from security experts for its strict controls and minimal data retention.
Starting with iOS 27—and already implemented in select iOS 26 apps—Apple is partnering with Google to power additional AI models. This collaboration extends PCC protections to Google Cloud infrastructure while maintaining strong security standards.
How the New Permission Prompt Works
The popup appears when users engage with specific AI-powered tools, such as:
• Shape generation features in iWork apps on iOS 26
• Similar generative tools in Freeform on iOS 27
The prompt clearly informs users that their data (typically just the typed prompt or edited image) will be processed on Google Cloud servers. Key details include:
• Google is prohibited from using the data to train its models or retaining it long-term
• Users can grant permission for one-time use or “always allow”
• The process remains isolated and aligned with Apple’s privacy commitments
This transparency measure was spotted in the iOS 27 beta and recent Apple Creator Studio updates, which preview advanced AI tools slated for broader release.
Apple’s Stance on Security
According to Apple, PCC on Google Cloud incorporates equivalent architectural safeguards to its on-device and Apple-silicon implementations. These include isolated processing namespaces, short-lived inference software, and confidential virtual machines for key management.
The company positions this as an evolution rather than a compromise, allowing more powerful AI features while upholding user control and privacy.
The move comes as Apple continues to roll out its Foundation Models, blending on-device, Private Cloud, and now select third-party cloud processing for optimal performance and capability.
MacDailyNews Take: What do you think? Does the explicit permission prompt make you more comfortable with cloud-based AI, or would you prefer to stick with only on-device processing?
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!
Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.
