A Canadian couple’s devastating ordeal recently put an urgent spotlight on internet security: In March, the Brantford, The couple — who is in their 70s — received a pop-up warning on their laptop claiming their accounts were involved in crimes. The warning locked their screen and instructed them to call a number. For five months, scammers posing as government and police officials convinced them to move their entire life savings — over $1,010,000 — into gold bars and bitcoin ATMs under the disguise of protecting their assets. Despite a financial advisor’s warning that the scenario sounded like fraud, the couple obeyed every order, ending up penniless, and facing a tax bill for their depleted Registered Retirement Savings Plans (RRSPs).
Scams like this, often referred to as scareware, have tormented internet users for years. Typically, a fake antivirus alert or a threatening pop-up hijacks the browser, pushing users to call a fake support line or click dangerous links. Now, the world’s two most-used browsers — Microsoft Edge and Google Chrome — are deploying artificial intelligence, specifically LLMs (large language models) and machine learning, to block these attacks before they start.
Microsoft Edge began ramping up protection with a new scareware blocker feature, which started rolling out to the public in October 2025. This blocker is now on by default for most Windows and Mac users. At its core is a local computer vision AI model, trained on real-world scam pages, that instantly recognizes and shuts down deceptive pop-ups. When it detects scammy behavior — like forced full-screen takeovers and shrill audio alerts — Edge force-closes the offending window, stops audio, and warns the user. Testers found this automated protection often blocked scams hours or even days before global blocklists noticed them.
Edge’s most recent improvement, introduced with browser version 142, is a scareware sensor that immediately reports newly detected attacks back to Microsoft’s SmartScreen cloud even before users report them. Instead of waiting for dozens of user complaints before a dangerous page is blocked worldwide, Edge now provides live intelligence: As soon as a page triggers Edge’s local AI, it can be examined and blocked for everyone, reducing user exposure by more than half, compared to previous delay-prone systems.
Google Chrome has a similar approach. Earlier this year, Google rolled out Gemini Nano, its smallest on-device language model, to Chrome. Whenever a user visits a page, Chrome’s AI analyzes the content in real time, scanning for signs of tech support scams, scareware patterns, or attempts to hijack browser features such as full-screen APIs. Chrome’s AI doesn’t upload the user’s personal data — instead, all analysis runs locally, preserving privacy and reacting to scams immediately, not after-the-fact. The moment a scam is detected, Chrome blocks the site, pops up a warning, and gives users the option to escape or report a false positive.
This shift toward real-time, on-device threat detection marks a major leap over traditional blocklist approaches. Blocklists typically respond only after a scam has victimized multiple users — sometimes after hundreds or thousands of attacks. By harnessing AI and LLMs that see scams with the same eyes as the victim, browsers can now catch attacks on the first victim, not the tenth or hundredth.
Both Edge and Chrome are updating their models and reaction speed continuously. Microsoft allows enterprise administrators to manage scareware controls by policy, letting companies whitelist internal resources or adjust levels of protection. Google’s Chrome team says it will expand on-device AI intelligence to catch even more fraud types, such as fake delivery notifications and government imposter scams, with notifications already shipping to mobile users as well.
Here’s the thing, folks: Computer errors especially those that target seniors or less tech-savvy individuals sometimes do not appear to be scams. Globally, the Global Anti-Scam Alliance estimates over $1 trillion in scam-related losses in the last year. The criminals are using AI to create ever more convincing pop-ups and fake dialogues — so browser makers have decided to fight fire with fire.
So, it is very important that you keep Edge and Chrome on the latest versions, and make sure advanced protection features are enabled. In Edge, scareware blocker is located under Privacy, search, and services. In Chrome, the feature is activated automatically in Enhanced Protection mode.
With that… This hard-earned progress can’t bring back a lost nest egg, but it does mean fewer tragic stories like the Ontario couple’s will unfold. By moving protection inside the browser and empowering AI to act instantly, both Edge and Chrome are offering ordinary people their best fighting chance against the evolving, relentless world of online scareware.
When you work with the technology every day that means you know when to issue a PSA!