AINews wire, 3 h ago
Chrome to test screenshot-based AI scam detection for Enhanced Safe Browsing users
Reported by Digital Trends

Google is working on a new scam detection feature for Chrome that captures screenshots of webpages and classifies them with AI, according to Digital Trends. The report says the work was spotted in recent Chromium development, which points to browser-side screenshot capture and visual analysis. The aim is to catch deceptive sites that can slip past current checks, including new scam domains that are not yet on Google's lists.
Chrome already uses on-device and server-assisted AI to check pages for scams. The new step adds visual classification, which looks at how a page appears to a user instead of relying only on its address, code patterns or known threats. Scammers often set up short-lived domains that copy real layouts, so the approach targets that gap.
The feature would apply only to users who turn on Enhanced Safe Browsing, Chrome's stronger security setting. Standard Safe Browsing users would not have their pages analysed this way unless they switch to the higher mode. The outlet compares the idea with Windows Recall, which takes periodic screenshots, but it notes one key difference. Windows Recall stores its snapshots locally, while Chrome's feature would send them to Google's servers.
Digital Trends says the privacy trade-off is still unclear. No release date has been announced, and the report does not confirm which regions or Chrome versions will get the feature. The source also gives no details on how long screenshots would be kept or how Google would use them.
Enjoying this? Get the AI brief
Every Thursday: AI that matters for your home.
✓ Free · seven editions · one click to leave
Original reporting by the sources above. Summary by smart-home.space.