Republished on March 12th with new reports into Apple’s AI setbacks, with serious implications for the on-device tracking and privacy warnings hitting Android users.
Google has an awkward Android problem that has just been highlighted. The AI space race between Google and Samsung and between Android and iPhone has exposed long-standing issues that need to be addressed, issues buried within Android’s core.
Google continues to narrow the gap to iPhone, with Android 15 and the forthcoming Android 16 bringing security and privacy upgrades. And while Samsung owners have endured a frustrating wait for their upgrade, this is now imminent and it seems Android 16 will be handled differently. But there’s one awkward black hole that remains, hidden tracking that’s impossible to stop. This is bad news for all Android users, but especially for Galaxy and Pixel users buying handsets with “on-device privacy” built in.
The issue is the old versus the new. You will have noticed that AI announcements from both Google and Samsung stress on-device, privacy-preserving processing when touting new features. Google’s latest scam detection “protects your privacy by processing everything on-device,” and even its controversial new photo scanning “is done on-device and all of the images or specific results and warnings are private to the user.”
Samsung — which pushed hybrid AI before Google — goes further, “Galaxy’s approach to AI personalization,” it says, “from the launch of Samsung’s first AI phone to our continuous innovation of on-device AI, is to never compromise on privacy.”
That’s the new. But then we come to the old. As I reported earlier this month, a study published by Trinity College, Dublin has exposed Google’s decision to track Android phones through “cookies, identifiers and other data that Google silently stores on Android handsets,” through the default apps that are pre-installed. This happens despite there being “no consent sought for storing any of this data and no opt out.”
There’s a legal dimension to this tracking, which the research team says has been exposed for the first time in their report. This, they say, is a “wake-up call” for data regulators to “start properly protecting” users of Android phones.
Google did respond on this legal point, telling me “the researcher acknowledges in the report that they are not legally qualified, and we do not agree with their legal analysis. User privacy is a top priority for Android and we are committed to complying with all applicable privacy laws and regulations.”
But on the substance of the tracking, the Trinity team says it “informed Google of our findings, and delayed publication to allow them to respond. They gave a brief response, stating that they would not comment on the legal aspects (they were not asked to comment on these). They did not point out any errors or misstatements (which they were asked to comment on). They did not respond to our question about whether they planned to make any changes to the cookies etc stored by their software.”
Google told me “this report identifies a number of Google technologies and tools that underpin how we bring helpful products and services to our users.” There is nothing to suggest the decision to maintain this on-device tracking is to be revisited given these concerns; as things stand the claims remain unaddressed.
When you add to this the decisions to delay the deprecation of cookies in Chrome and equally controversially to resurrect digital fingerprinting, you start to see a pattern. Despite the focus on on-device AI processing, the basics have not changed. It suggests we may be seeing an AI privacy push that’s more marketing than privacy related.
This leaves Samsung in a strange limbo land. It is pushing to catch iPhone, but it’s tied to Android. And that means these kind of legacy tracking issues impact all its users. At the same time, it cannot rush out security fixes or OS upgrades at the same pace as Pixel, because it does not control both its hardware and core software.
Meantime, while there remains a security and privacy gap between Android and iPhone, that has been reversed on the AI front. The iPhone-maker finds itself in a mess having failed to meet the expectations it set for 2025’s Apple Intelligence upgrade, which was central to the launch of the iPhone 16 and latest iPad Pro. And that will take the focus away from the security and privacy differences and new warnings.
As Google promotes its ever advancing on-device AI assistant, Apple’s looks ever more dated. “Siri isn’t an assistant, it’s an embarrassment,” Macworld has just complained. “Apple’s inability to make its signature assistant good, let alone competitive, should have heads rolling in Cupertino.” This follows Apple admitting that “the big new Siri features that might have made Siri pretty good have been delayed for a year (maybe more). In the meantime, we don’t see any reason to believe that Siri won’t continue its pitiful pace of ‘improvement’ that has left it miles behind its AI competitors.”
Apple made AI privacy with on-device processing and its secure cloud enclave the central theme of its offering, a differentiator to Google and Samsung. With Apple yet to make this work, there is no viable, more secure alternative to the Google-led Android AI. This matters because while Google and Samsung are promoting on-device AI processing, the dividing line with off-device, cloud processing is still blurred behind various privacy policies.
There is also speculation that security is behind the delays to Apple’s AI upgrade, taking pressure off competing offerings. As 9to5Mac says, “it does seem entirely plausible that security is also a key factor. Apple has turned privacy into a key differentiator and selling point for its products, so any vulnerability that allows a rogue app to access and export your personal data would be a complete disaster for the iPhone maker.”
And so here we are instead. While this remains fresh news, it does have the potential to turn the Android versus iPhone debate on its head. The longer there’s no viable AI alternative on iPhone to Google’s full Gemini advances, the more the focus falls away from security and privacy given the feature differentiation.
None of this changes the fact we need transparency and clarity as to what is being done on-device, how and by whom. It would be good to see a thorough review of the hidden tracking taking place without consent or opt out — that means “silent” Android tracking and digital fingerprinting. It would be good to see default opt outs introduced across the board. Absent that, each of those on-device marketing messages may need an asterisk.
