Every service we look at, by type of protection
A service is on this page when its own site says what it does about AI risks, and we can quote that with a link to the page and the date we read it. Where a provider names no AI feature, the row says so instead of filling the gap for it.
This is a signpost, not a ranking: inside an area the list is alphabetical, without scores. How we judge and price a service is on the methodology page.
Narrow the list
20 of 20 rows shown
AI scams and deepfakes
Read more about this areaA call or message that seems real and asks for money in a hurry.
Avast
AI scams
Avast Assistant read on September 22, 2026
You show it a suspicious message or offer and it says whether it looks like a scam.
Does not do: Automatic email screening starts at Avast Premium Security and is not in the free product.
Free to use
From $4.09/mo
price read on September 22, 2026
Compare Avast vs Malwarebytes: Assistant or Scam Guard?Bitdefender
AI scams
Scamio read on September 23, 2026
Rates a message, link or QR code that you paste into the chat yourself.
Does not do: Scamio checks only what you hand it, so it does not scan your device or your inbox.
Free to use
F-Secure
AI scams
AI-powered SMS protection read on September 22, 2026
Checks incoming texts from senders outside your contacts and moves scam messages to your junk folder.
Does not do: The AI filter reads only text messages from senders outside your contacts, on iOS and Android. WhatsApp and other chat apps are not named.
Where it works: Offered in the Netherlands. F-Secure runs Dutch pages for Total and for Scam Protection.
F-Secure Total, Internet Security and the standalone Scam Protection
From $4.99/mo
price read on September 22, 2026
Compare F-Secure vs Bitdefender: Deepfake Detector or Scamio?Guardio
AI scams
Predictive threat detection read on September 22, 2026
Blocks scam sites and phishing links in your browser before you click.
Does not do: On a computer Guardio works only as an extension in Chromium browsers. Mail checking covers Gmail only.
all paid plans
From $9.99/mo
price read on September 22, 2026
Compare Malwarebytes vs Guardio: Scam Guard or blocking?Malwarebytes
AI scams
Malwarebytes Scam Guard read on September 22, 2026
Judges a text, number or screenshot that you share in the app.
Does not do: Scam Guard reads only what you share with it, so it does not watch your messages by itself.
in Malwarebytes Premium Security
NordVPN
AI scams
Scam, phishing, and malware protection read on September 22, 2026
Blocks scam sites and harmful links in email and scans your downloads for malware.
Does not do: The AI-backed detection runs only on Windows and macOS. Android and iOS get the version without nexos.ai.
from Complete
Norton
AI scams
Norton Genie read on September 22, 2026
You hand Genie a text, email or link and it says whether it looks like a scam.
Does not do: Deepfake Protection only covers video in English on supported platforms.
all paid plans
Surfshark One
AI scamsLeaked data
Email scam checker read on September 22, 2026
Checks an open Gmail message for phishing when you click to scan.
Does not do: The email check works only in the Chrome extension on Gmail.com. It runs when you click to scan, not over your whole inbox.
from Surfshark One
TotalAV
AI scams
No AI feature of its own
TotalAV names no AI feature of its own. WebShield blocks fake and phishing websites, and every paid plan lists Phishing Scam Protection.
Leaked data and account takeover
Read more about this areaYour details are out there and someone could get into your accounts.
Dashlane
Leaked data
Scam Protection read on September 22, 2026
An AI model checks every website in your browser and warns you before you type details into a fake page.
Does not do: Works only on a computer in Chrome, Firefox, Edge and other Chromium browsers. Dashlane states that scam alerts are not available on desktop Safari or on mobile devices.
Where it works: Offered in the Netherlands. The Dutch pricing page lists it under Premium as "Nieuw! Stop oplichting in realtime". Whether the warning itself appears in Dutch is not measured.
from Premium (also in Friends & Family)
DeleteMe
Leaked data
AI feature not recorded yet
US data-removal service; enrollment is limited to United States residents.
Incogni
Leaked data
Auto-removal from 420+ brokers read on September 22, 2026
Sends removal requests to data brokers again and again, so scammers can buy less about you.
Does not do: Incogni checks no messages, links or calls and spots no cloned voice.
Where it works: Available across the EU, though the published broker list is mostly American.
all paid plans
NordPass
Leaked data
No AI feature of its own
NordPass has no AI feature of its own. It offers its Data Breach Scanner and Email Masking as the defence against AI that guesses passwords.
Optery
Leaked data
AI feature not recorded yet
Removes your data from broker sites; not available for Dutch or EU addresses.
From $3.25/mo
price read on September 22, 2026
Compare Incogni vs Optery: auto-removal or AI Generated ReportsProton Pass
Leaked data
Proton Sentinel read on September 22, 2026
AI systems and human analysts watch sign-ins to your Proton account and stop the suspicious ones.
Does not do: Sentinel guards your Proton account and not the messages, links or sites you meet elsewhere.
from Pass Plus, also in Pass Family and Proton Unlimited
Surfshark One
AI scamsLeaked data
Email scam checker read on September 22, 2026
Checks an open Gmail message for phishing when you click to scan.
Does not do: The email check works only in the Chrome extension on Gmail.com. It runs when you click to scan, not over your whole inbox.
from Surfshark One
Keep your data out of AI
Read more about this areaWhat you write or store could end up training someone else’s AI.
Gmail
Data out of AI
AI feature not recorded yet
Google's free inbox, with an AI setting worth switching off.
From €1.99/mo
price read on September 22, 2026
Compare Proton Mail vs Gmail: Proton Scribe or smart featuresNordLayer
Data out of AI
Clipboard control in NordLayer Browser read on September 22, 2026
An admin can stop staff pasting text into an AI chat inside the NordLayer Browser.
Does not do: The rules work only inside the NordLayer Browser. Nothing on the pages read says they apply in Chrome, in Edge or in the ChatGPT app.
Where it works: Not stated. NordLayer has no Dutch pages and charges in USD with local tax added.
from NordLayer Browser Premium
From $4/mo
price read on September 22, 2026
No comparison yet
Proton Drive
Data out of AI
Proton Docs read on September 22, 2026
An editor inside Proton Drive that keeps your text encrypted, so Proton cannot read or train on it.
Does not do: Proton Drive has no AI that spots a scam. The protection is encryption, not detection.
Free to use
Proton Mail
Data out of AI
Proton Scribe read on September 22, 2026
An AI writing assistant inside Proton Mail that drafts and improves email without training on your inbox.
Does not do: Scribe writes email. It does not check incoming mail for phishing or scams.
Included with Proton Duo, Family, Visionary and Lifetime. A paid add-on on other paid plans, with a free trial of 14 days. Not in Proton Free.
Why isn’t McAfee here?
- McAfee
- McAfee names Scam Detector and Deepfake Detector, but Deepfake Detector runs only on selected Copilot+ PCs in a handful of countries. source, read on September 22, 2026
- AVG
- AVG names artificial intelligence for malware detection on its own page, but no AI feature against scams or deepfakes. source, read on September 23, 2026
- Avira
- Avira names AI-powered analysis against malware on its own pages and no AI feature against scams or deepfakes. source, read on September 22, 2026
- ESET
- ESET writes about advanced artificial intelligence against malware and phishing, but gives that protection no feature name. source, read on September 23, 2026
- Kaspersky
- Kaspersky writes a lot about AI scams on its blog, but its Premium page names no AI feature. source, read on September 23, 2026
Where we earn from a click, it says so under the button.