Google Alerts Users to New AI Scam Targeting Gemini Users
Google warns users about a scam exploiting its Gemini AI system, prompting enhanced security measures.
Key Points
- • Google alerts 1.8 billion users about a new AI scam exploiting Gemini assistant.
- • The scam uses hidden email commands to extract sensitive data without consent.
- • Traditional security measures may not detect this stealthy method of attack.
- • Google is enhancing security protocols and advises users to be cautious with emails.
Google has warned its 1.8 billion users, particularly impacting travelers, about a sophisticated new cyber scam that exploits its AI assistant, Gemini. This alarm comes as the scam employs hidden commands within emails to extract sensitive information like passwords and financial details without users’ awareness. Unlike conventional phishing methods that necessitate users to click on malicious links, this scam operates at the linguistic level, making it more stealthy and harder to detect.
Google categorizes the threat as a transition from more direct scams to one where machines are manipulated through interactions with users. In light of this, the company is enhancing its security measures to better identify and mitigate such indirect prompt injections. As part of its guidance, Google urges users to exercise caution with unfamiliar emails and interactions with AI, alongside enabling two-factor authentication for added security.
This evolving threat underscores a significant shift in the landscape of cyberattacks, highlighting the need for robust protective frameworks as AI becomes increasingly integrated into everyday tools.