Chrome adds pre-visit warnings for sites it deems suspicious

Google is introducing a new safety feature in Chrome that warns users about potentially risky websites before they open them, according to The Verge. The change creates a graded response to online threats, sitting between an ordinary page load and the full-screen block the browser already displays for sites it considers actively harmful.

How the two warning tiers differ

The new notices apply to websites that share characteristics with pages previously linked to phishing, malware and scams. According to The Verge, Google describes these flags as carrying a "lower confidence verdict" — meaning a mistake is possible — though users are advised to continue only if they are fully certain a site is legitimate. A verdict of "dangerous" sits at the opposite end of the scale: it reflects higher confidence and a high risk of harm, and Google's guidance is to avoid such sites entirely.

What the warnings look like

The two categories are designed to be told apart at a glance. A suspicious-site alert appears as a pop-up window on both desktop and mobile, with the underlying page left visible but darkened, and it includes a "Mark as safe" control for cases where a legitimate site has been flagged by mistake. The warning for dangerous sites is far harder to ignore: The Verge reports that it takes up almost the entire screen and replaces the page with a solid red background.

Built on an existing model

The approach mirrors a two-stage system Google introduced for Chrome file downloads in 2024, which applied visually distinct notifications to blocked files depending on whether they were judged suspicious or dangerous. The site warnings extend that same confidence-based logic to navigation.

Alongside the redesign, Google says it has improved its Gemini-powered scam detection for users with Enhanced Protection enabled. One question remains open: The Verge reports it is unclear whether the suspicious-site notifications are available to all Chrome users or only to those on Enhanced Protection, and says it has asked Google to clarify.

Why it matters

Chrome's response to a malicious site has so far been blunt: a red, near-full-screen wall. The new tier gives the browser a way to express doubt, acknowledging that threat detection is probabilistic. Blocking every borderline site outright would lock people out of legitimate resources, while ignoring low-confidence signals would leave users exposed. The dimmed-page presentation and the "Mark as safe" option are the compromise — an interruption strong enough to prompt a second thought before someone enters credentials or payment details, but reversible when the algorithm gets it wrong.

For website operators, the flip side is a new misclassification surface: a legitimate site that happens to share traits with scam pages can now lose visitors at the doorstep, with no recourse beyond hoping users click through. For everyday users, the feature adds an earlier, gentler signal in exactly the moment — right before a page loads — where phishing attacks do their damage. The rollout also shows how warning design itself, not just detection accuracy, is becoming a core part of browser security, and how Google is weaving its Gemini models into Chrome's protective layer for those who opt into its strongest settings.