Apple’s New Call Screening Feature: How it Works, Why it Will Cut Spam Calls, but Won’t Protect Against Scams
- Richard Gotlieb
- Oct 14
- 5 min read
Updated: Nov 19
Intro
As a devout Apple supporter, I found a hack to get my iPhone to switch to Apple’s new iOS26, and have to say that despite the fact that it drains your battery faster, and has added a slight lag in my phone, overall, I was impressed. In my mind, the most interesting feature is Apple’s new Call Screening feature which automatically answers unknown numbers, asks the caller for their name and reason for calling, transcribes the reply in real time, and shows it to you so you can decide whether to pick up. For the deluge of spam and nuisance calls this is huge. Unfortunately, for sophisticated social engineering, it is negligible.
How the feature actually works
When a phone number that is not in your contact list rings your phone, iOS can answer in the background. A Siri style voice prompts the caller to state their name and purpose.
The caller’s reply is transcribed and shown on your screen in speech bubble form, and then you can accept, decline, or send a message. You can also choose to silence unknown callers and rout them to voicemail. This feature allows users to introduce a level of friction between them and unknown callers.
Why this will cut spam calls
Spam and robocalls are mostly high volume, low effort. Automated systems and lazy spammers cannot reliably answer the feature’s free form prompt in a way that would get the user to accept the call. Add a tiny hurdle and many automated networks simply fail, or produce some mangled response that users would be able to recognize as not a call they need to pay attention to. Early reporting and user testing show substantial reductions in nuisance calls when people turn the feature on. If your problem is constant interruptions and robo noise, this is an immediate win.
Cuts the calls, but not the scams
Scammers are not spambots. Scammers are social engineers who adapt. They are already answering the prompt and still getting through. They will say a plausible line like, “Delivery for Mary Jones,” or “We are calling from Your Bank about suspicious activity.” That gets them through, and then the real attack begins: pressure, urgency, deception, requests for money, account credentials, or remote access. In case you’re wondering how a scammer knows your name, or the name of your bank, you would be frightened to learn how much information they may already have at their fingertips. Things like your name, phone number and address are already on the internet. Sites like Radaris, Whitepages, TruePeopleSearch, and Spokeo scrape property deeds, voting rolls, old phone directories, and whatever data brokers will sell them. Results are hit-or-miss, but for a scammer to have a great day, all it takes is one scam to work. Scammers also have access to databases that scour the obituaries everyday, that can add up to an entire family tree, complete with granddaughter and grandson names. Periodic data leaks provide scammers with not just your personal information, but often include things like your credit card and the name of your bank. Aside from the breaches at banks or credit reporting agencies, there are dozens of other sites that have bits and pieces of your core information. To check if any companies you've used have experienced data breaches, try https://haveibeenpwned.com/. This free service lets you enter your email address to see if it has been compromised in any of the data breaches it tracks.
Put bluntly, Apple’s voice prompt does not verify identity, intent, or the content of a conversation. Call screen only verifies that a human is on the other end of the line For conversational scams you need real time reasoning about behavior and persuasion patterns, not a gate that asks for a sentence and moves on.
Other blind spots and edge cases
Spoofing and number trickery. If a scammer spoofs a trusted number like your child’s or mother’s phone number (discovered by one of the sites named above) or uses a lookalike number, call screening does not verify the identity behind the call.
Legitimate callers get confused. Delivery drivers, techs, and businesses may fail the prompt and never get through. That creates friction and frustration for real people. For example, I was trying to get in touch with customer service from Amazon, and after asking for a call through their website, they were going to call me back. Never heard from them, and after requesting another call back, and not hearing from them, it occurred to me the Call Screening may may have tripped them up. I disabled Call Screening, requested another call back, and sure enough received the call from Amazon a few minutes later. If you are expecting an important call, you might want to disable the Call Screening Feature.
Farming social trust. A carefully run scam can use short, plausible answers to pass the screen and then proceed to build rapport. The prompt does not block follow up persuasion tactics.
False reassurance. People may assume that a screened call is safe. That cognitive bias will be the thing scammers exploit the most.
What would actually protect people from scams
Stopping scams requires multiple layers that reason about context, content, and intent in real time:
Conversational detection. Specialized AI Models trained on social engineering s scams that look for persuasion patterns, identity impersonation, and ask for money or sensitive information. This is what OnGuardAI has built and approach the industry is moving toward.
Trusted contact escalation. If a call looks suspicious, notify a preselected contact for another set of eyes and ears. Social engineering is about isolation. Reducing that isolation reduces scam success rates.
Practical advice for users
Turn Call Screening on if robocalls are your main nuisance. It will quiet the phone.
Never assume a screened call is safe. If a caller then asks for money, gift cards, or remote access, hang up and call the institution directly using a number you trust.
Use layered protections. Carrier spam filters, specialized AI detection apps, verified caller systems, and a trusted contact plan together meaningfully reduce risk.
Bottom line
Apple’s Call Screening is a needed improvement to daily life. It reduces noise and weeds out lazy spam. It does not stop scammers who speak, think, and improvise. If your goal is to stop financial loss, look for apps that employ specialized AI Models that analyze conversations. Call Screening is a good tool, but does not fix the lock on the front door.