The scenario is always the same. A fake bank advisor on the line, with an urgent tone telling you to confirm quickly, and you validate a transfer you'll deeply regret before the end of the day. That's precisely the mechanism Apple wants to disrupt with Trust Insights, a feature of iOS 27 quietly unveiled at WWDC in June - the company's big developer conference where it previews upcoming systems - and detailed in a technical session called "Meet Trust Insights."

The goal is actually fairly straightforward: give apps a way to detect, while you're using them, that someone is probably pulling a fast one on you. Not an alert after the fact, once the money has been sent to the other side of the world. A real-time alert, right in the moment.

Trust Insights is an API - a toolkit Apple makes available to developers so they can plug this detection directly into their own apps, without having to build it themselves. Your bank, your payment app, or your messaging app can pick up the feature and integrate it into their sensitive screens, provided they've been granted a special authorization from Apple.

To sniff out trouble, the system cross-references your interaction patterns, the rhythm of your gestures, the context of the action, and a few basic sensor readings. Someone dictating what to type while you're on the phone shows up in the way you go about it: hesitations that don't match your usual behavior, abnormal back-and-forth between screens, a typing cadence that's off compared to normal.

On the privacy side, Apple is adamant: the content of your Messages, Mail, and Photos is never inspected. All processing happens on the iPhone, the raw data is discarded immediately, and only one piece of information ever leaves the device - a risk level of medium or high. Nothing more.

It's then up to the app to decide what to do with it. It can display a prominent warning, impose a delay before allowing a transfer to go through, or require an additional verification step. The kind of forced pause that gives you ten seconds to hang up and ask yourself why this "police officer" is demanding immediate payment.

Trust Insights organizes the actions worth monitoring into five broad categories: payments and money transfers, account modifications, access to costly resources, sending messages or signing documents, and a catch-all category for everything else. In short, all the moments where a scammer has the most interest in holding your hand right to the end of the operation.

Targeted scams are the most problematic: fake tech support, impersonation of authority figures, a fake cop or banker pressuring you to act without thinking - and even harder to handle, the fake family emergency. These scams are as old as the telephone itself, but they've been upgraded with deepfakes, those AI-generated voices and videos capable of convincingly imitating someone you know.

The catch, of course, is that everything depends on developers' goodwill. If your bank doesn't bother integrating Trust Insights into its app, absolutely nothing will happen. And you'll have to wait until fall anyway, when iOS 27 launches publicly, to see the first integrations show up on your screen.

That said, the idea is a pretty good one. A ten-second forced pause at the right moment will protect more bank accounts than years of awareness campaigns ever could.