Google says its AI found 500+ security holes in its own websites
Google said an internal AI tool found more than 500 website weaknesses in its own systems that could have let attackers make pages behave in harmful ways inside a visitor’s browser. The main takeaway for the public is that even major online services can have hidden website flaws, so users should stay alert for unusual behavior when using web apps.
Who is at risk
People who regularly use online accounts, web apps, and browser-based services are most at risk if a trusted site is compromised.
What to watch for
Watch for unexpected pop-ups, surprise sign-in requests, pages that suddenly redirect, or messages asking you to click, approve, or re-enter account details on a site you already use.
What to do
Keep your browser updated, turn on two-step verification for important accounts, and avoid clicking unexpected prompts or re-entering passwords unless you manually navigate to the site yourself.