iOS 27 to detect when user may be getting scammed
What's the story
Apple has a new framework in its upcoming iOS 27 update, called Trust Insights. The innovative tool is designed to help apps detect when a user might be falling victim to social engineering scams. These scams often occur through voice calls, text messages, emails, and more. Trust Insights works by analyzing interaction patterns, timing, and context, as well as basic sensor data on-device.
Scam detection
Addressing the growing threat of AI deepfakes and scams
The introduction of Trust Insights comes as a response to the growing prevalence of tech support scams, authority impersonation, and family emergency fraud. These types of scams have become more common in recent years, especially with the rise of AI deepfakes. The framework works by detecting signs that a user might be getting coached into a scam and assigning it a medium or high risk level.
User protection
How Trust Insights works
Once a potential scam is detected, Trust Insights allows the app to take preventive measures. These could include issuing warnings, introducing delays, or requiring extra verification steps. Apple has clarified that Trust Insights doesn't look at the content of Photos, Messages, or Mail. Instead, it focuses on behavioral signals on-device and immediately discards the underlying data after sending just one value to Apple's servers.
User control
Users can disable the feature
While users can disable Trust Insights in Settings, Apple has warned there could be a cooldown period. This is to prevent users who may have been coached into turning it off from doing so. The framework will initially cover five main operation categories: payment, account updates, resource use requests, communication, and other operations that don't fit the above categories.