How to Limit What Apple’s New Siri AI Can Access in iOS 27
Original reporting by Electronic Frontier Foundation
Apple's new Siri AI refers to a significantly re-engineered version of its voice assistant, now deeply integrated with iOS and behaving more like a comprehensive chatbot. This powerful update transforms the familiar blob-like assistant into a robust application, unifying with Spotlight search to provide expansive access across your device's apps and content. The goal is unprecedented utility, enabling Siri to contextually understand requests, summarize conversations, and perform complex tasks by drawing directly from your personal data—whether it's emails, messages, or third-party app content.
Balancing utility and privacy
However, this enhanced capability comes with a critical privacy trade-off: a blurred line between on-device processing and data sent to Apple’s Private Cloud Compute (PCC). Users are often left without clear indication of when their data leaves their device, fundamentally altering the risk assessment, especially for sensitive information or those utilizing advanced encryption. While Apple maintains that data sent to PCC is not stored post-processing, default settings grant Siri AI broad access to personal communications and app content. A new "on-screen awareness" feature can even transmit data from secure chat apps to the cloud without explicit user or developer intervention. Navigating this new landscape requires careful attention, but users are not entirely without recourse; several granular settings are available to restrict Siri AI’s reach and manage their digital privacy, offering a measure of control over its newfound power.
Apple's integration of AI into Siri marks a significant leap in personal assistant capabilities, offering unprecedented utility by deeply engaging with user data. However, this advancement is inextricably linked to a complex web of privacy trade-offs and a challenging user experience for managing data access. While Apple has provided some mechanisms for control, the inherent ambiguity surrounding on-device versus cloud processing, coupled with the granular and often hidden nature of privacy settings, places a considerable burden on users to safeguard their information. The "on-screen awareness" feature, in particular, highlights a critical gap where user agency and developer control are currently insufficient.
The Path Forward Beyond the immediate concerns, Siri AI's evolution carries profound implications for the future of personal technology. Apple, a long-standing advocate for user privacy, is now navigating the intricate balance of powerful AI and data protection. Its choices will undoubtedly set a precedent for the entire industry, influencing how other tech giants develop and deploy their own AI assistants. This move also fundamentally redefines the relationship between users, their devices, and the data they generate, shifting expectations for what an "intelligent" device can and should do. The ongoing challenge for Apple, and indeed for the broader tech ecosystem, will be to empower users with genuinely transparent and intuitive controls, ensuring that the convenience of AI does not erode the foundational principles of data privacy. The demand for clear, unified privacy toggles and explicit indications of data handling will only intensify as AI becomes more deeply embedded in our digital lives.
Frequently asked questions
- What is new about Apple's Siri AI and how does it differ from the classic version?
- The new Siri AI is a significant upgrade, resembling an AI chatbot rather than a traditional voice assistant. It's deeply integrated with Spotlight, unifying search and assistance into a single interface. This version offers enhanced capabilities to surface details from your device, acting more like an app with expanded access to your content, making it more powerful and context-aware than its predecessor.
- How can users manage privacy settings and control data access for the new Siri AI?
- Users can restrict Siri AI's access by disabling "Show Content in Search" for individual apps in Settings, preventing it from indexing their content. It's also possible to opt out of using your data for AI training under Privacy & Security settings. However, a feature called "on-screen awareness" can send displayed content to Apple's cloud without an immediate user control to block it for specific apps.
- What are the main privacy considerations when using the new Apple Siri AI features?
- A key privacy consideration is how data is processed: some on-device, some sent to Apple’s Private Cloud Compute (PCC) without clear user indication. While Apple states PCC data isn't stored post-processing, sending end-to-end encrypted data off-device changes its risk profile. The "on-screen awareness" feature can transmit currently displayed content, like secure chat conversations, to PCC, with limited user or developer control.