Apple Expands AI on iPhones: What Does This Mean for Your Privacy?
Amsterdam, maandag, 11 augustus 2025.
Apple has recently announced Apple Intelligence, a new AI platform that will be installed by default on iPhones. Although Apple has always been committed to privacy, this step raises questions about the balance between convenience and data protection. Recent research by Lumia Security has shown that Siri, Apple’s AI assistant, sends more sensitive user data to Apple servers than necessary, even when users have adjusted their privacy settings. This development calls for careful consideration between the benefits of AI and the protection of personal information.
Apple Intelligence: A New Era for iPhones
Apple recently announced the introduction of Apple Intelligence, its own AI platform that will be installed by default on iPhones. With over a billion Apple users worldwide, this step will significantly increase the availability of AI. According to Laurens Verhagen, an AI journalist at the Volkskrant, this marks a new phase in the availability of AI technology [1]. Apple Intelligence processes calculations on the device itself, rather than on external servers, which is a significant difference from competitors such as Google and OpenAI [1].
Privacy as a Core Value
Privacy is a core value for Apple, which has long been committed to protecting user data. According to Verhagen, privacy is where Apple has always positioned itself. Calculations on iPhones and iPads take place on the device itself, not on external servers, which should enhance data security [1]. However, recent research by Lumia Security has shown that Siri, Apple’s AI assistant, sends more sensitive user data to Apple servers than necessary, even when users have adjusted their privacy settings [2].
Data Collection by Siri
According to research by Lumia Security, presented at the Black Hat USA conference, Siri sends the content of dictated messages and commands, including WhatsApp communications, to Apple servers, even when this is not necessary to execute user requests. Siri also automatically scans users’ devices for installed applications related to voice queries and sends this information along with location data to Apple servers [2]. This can cause privacy issues, especially since WhatsApp messages are relayed through Apple’s Private Cloud Compute infrastructure, potentially compromising end-to-end encryption [2].
Apple’s Response
Apple acknowledged the issues in February 2025 but claimed in July 2025 that this behaviour is not a privacy problem. The company states that existing privacy policies are sufficient to cover these practices. However, Yoav Magid, a senior security researcher at Lumia Security, argues that this is not enough and that more transparency and control are needed [2].
Balancing Convenience and Privacy
The integration of AI technology certainly offers many benefits. For example, users can draft emails with a single button press, which many find very convenient [1]. However, a careful balance must be struck between the conveniences offered by AI and the protection of personal data. The option to refuse permission for AI use is available, but it is becoming increasingly difficult [1].
Security Measures and Recommendations
To reduce privacy risks, experts recommend reading Apple’s privacy policies, implementing network-level monitoring, and demanding transparency from AI providers [5]. Magid also advises blocking network traffic to guzzoni.apple.com and disabling unnecessary app settings [5].