Apple details opt-in iOS 27 scam-detection feature
Impersonation Risk Detection returns an Unknown, Medium or High risk level to apps, and Apple says it never receives the data behind it.
Apple has detailed Impersonation Risk Detection, an opt-in feature in iOS 27 and iPadOS 27 that checks for signs of social engineering scams before sensitive actions such as payments, password changes or sharing account details, according to ZDNET.
The feature returns one of three risk levels to the requesting app: Unknown, Medium or High. The app, not Apple, decides how to respond to a high level — for example by asking for identity verification, showing a warning or delaying the action.
Impersonation Risk Detection is off by default. To enable it, users go to Settings, then Privacy & Security, then Impersonation Risk Detection, and turn on "Share with App Developers". The setting can take up to 24 hours to take effect.
The Privacy & Security screen lists participating apps and lets users turn off access for individual apps. Apple says it does not receive the data used to generate the risk level and does not analyze content in Mail, Messages or Photos.
Both built-in and third-party apps can support the feature, though Apple's support page lists no specific apps.
Quick answers
Is Impersonation Risk Detection on by default in iOS 27?
No. The feature is off by default and must be switched on in Settings under Privacy & Security.
What risk levels does Impersonation Risk Detection return?
It returns one of three levels to the requesting app: Unknown, Medium or High. The app decides how to respond to a high level.
Does Apple receive the data used to calculate the risk level?
Apple says it does not receive the data used to generate the risk level and does not analyze content in Mail, Messages or Photos.