AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Your iPhone can spot scams for you with iOS 27 – but you have to opt in

Collected Oct 1, 2026

Apple has shipped a security feature called Impersonation Risk Detection in iOS 27 and iPadOS 27, aimed at social engineering scams that get past conventional security checkpoints by persuading a person to make a risky choice. The feature is not enabled out of the box. Users must switch on a "Share with App Developers" option in Settings under Privacy & Security, and Apple notes the change may take up to 24 hours to take effect.

The premise is that many damaging scams do not defeat a password or a second factor so much as they convince the account holder to hand over control. The feature is designed to intervene at the moment a user is about to do something consequential inside a supported app, such as making a payment, changing a password or sharing account details. When the app requests an assessment, Apple analyzes the user's Apple account and information on the device for signals of a scam, then returns a risk level to the app.

There are three levels. Unknown means nothing suspicious was found but does not certify the action as safe. Medium indicates some signs of suspicious activity. High means significant signs were detected. The app, not Apple, decides how to respond; at a high level it might ask the user to verify identity, display a warning or delay the action.

The privacy design is the notable part. Apple says it never receives the underlying data used to build the risk level, and it does not analyze content in apps such as Mail, Messages and Photos. It does learn the type of action attempted when an app asks for an assessment. Participating apps receive only the risk level, not the inputs behind it.

The opt-in default is explained by access: a supported app needs a certain level of visibility into user actions and activities for the system to work. The same settings screen lists every app taking part, shows which apps requested an assessment and which actions triggered it, and lets the user revoke access for individual apps. Both built-in and third-party apps can support the feature, though Apple's support page names no specific apps.

For developers, this reads as another OS-level signal to fold into risk decisions rather than a replacement for existing fraud checks. In practice, an app that receives a high rating has to choose an appropriate friction: step-up verification, a warning interstitial or a short delay. A likely trade-off is friction on legitimate transactions versus catching scams that authentication alone will not stop, and the effectiveness will depend heavily on how many apps integrate the signal and how they act on it. Questions remain about false positives and how apps will treat the ambiguous Unknown level, but those are open points rather than stated results.

Why it matters: product and engineering teams building payments, account management or messaging features on iPhone and iPad gain a system-provided risk signal they can wire into sensitive flows, but only if users opt in and only for apps that implement it. The 24-hour activation delay and revocable per-app access suggest adoption and testing will be gradual, and teams should plan how their UX handles each of the three levels before shipping support.

Read at ZDNet · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Impersonation Risk Detection tries to spot scams traditional security methods might miss.