Google Gemini Vulnerability Exploited via Malicious Android Notifications
Researchers discovered that malicious notifications from WhatsApp, Slack, and SMS could exploit a prompt injection vulnerability in Google Gemini on Android, potentially allowing attackers to hijack the AI system. Google patched the issue server-side in November.
What happened
Researchers at SafeBreach discovered that malicious notifications from WhatsApp, Slack, and SMS could exploit a prompt injection vulnerability in Google Gemini on Android devices. The notifications could contain hidden commands mixed with benign text that Gemini would process without the user needing to open or tap on the message.
Reported impact
The vulnerability could potentially allow attackers to hijack Google Gemini and make it work on their behalf, affecting millions of Android users.
Index analysis
The flaw enabled prompt injection attacks through notification processing, a vector not previously associated with communications platforms on Android devices.
Governance controls
Google patched the issue server-side in November.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.