Apple has introduced Impersonation Risk Detection in iOS 27 and iPadOS 27, a security feature aimed at social engineering scams that trick users into granting access themselves. When someone takes a sensitive action, such as making a payment or changing account security details, the system examines device behavior in real time to spot signs of manipulation.
The analysis happens entirely on-device, so Apple does not see photos, texts, or other personal content. A supporting app that requests a risk assessment receives only a label: Unknown, Medium, or High. The app developer then decides what to do next, such as asking for identity verification, imposing a short waiting period, or displaying a warning.
The feature is opt-in and can be enabled in Settings > Privacy & Security. Apple warns that if someone contacts you and insists you turn it off, that is a red flag; the toggle can take up to 24 hours to take effect. It only works with apps built to support it, and Apple has not yet published a list of those apps.