Skip to playerSkip to main content
  • 2 days ago

Category

People
Transcript
00:00Google sends urgent warning to its 1.8 billion users.
00:05Google warns Gmail users about a phishing trick, exploiting Gemini's AI.
00:10Attackers embed hidden commands in emails, triggering fake security alerts, urging calls
00:16to extract personal info like passwords.
00:20Researchers exposed this in July.
00:25This scam uses indirect prompt injection targeting Gmail's AI summary function.
00:31Text, hidden by color, instructs the summarizer to fake warnings like, call support now.
00:37Your password was stolen.
00:39This type of scam is very hard to detect as there are no suspicious links to click on
00:45and the warning seems to come from Gmail itself.
00:48Assumed safety leads users to disclose vital personal information.
00:54It feels legitimate because it appears directly inside the summary window.
01:01Google says Gemini Workspace uses layered defenses to detect and stop malicious prompts.
01:07Treat suspect messages cautiously.
01:13As we have learned, never act on security alerts in summaries.
01:17Check myaccount.google.com for real Google activity.
01:25Phishing techniques may change, yet they still pressure quick action to steal details.
01:30Google's AI systems being manipulated highlights risks.
01:35Treat AI outputs cautiously, using them for convenience, not trust.
01:41To stay safe, remain skeptical, and verify threats.
01:45Open worrying alerts, checking grammar and senders, never share phone passwords, use Gmail
01:51help for assistance, and report phishing to enhance online safety.
Be the first to comment
Add your comment

Recommended