Google Gemini AI Bypass Lets Attackers Send SMS/WhatsApp from Locked Android 16 Devices
What Happened – A newly disclosed vulnerability in Google’s Gemini AI assistant on Android 16 permits an attacker with physical access to a locked phone to send SMS or WhatsApp messages without entering the device PIN. The exploit relies on a specific multi‑touch gesture that tricks the lock‑screen prompt into bypassing authentication.
Why It Matters for Compliance & Audit Readiness
- Demonstrates a gap in access‑control enforcement that SOC 2 CC 6.1 (Logical Access) expects organizations to monitor and evidence.
- Physical‑device handling is a control area often omitted from continuous‑compliance evidence; this incident shows why it must be tracked.
- Verisq’s SOC 2 Access Controls capability provides continuous monitoring of lock‑screen policy changes and immutable audit trails to prove that only authorized users can invoke privileged functions.
Who Is Affected – Consumer Android users, enterprises with BYOD or mobile‑device‑management (MDM) programs, and any organization that relies on Google’s Gemini for productivity.
Recommended Actions –
- Immediately enforce a policy that disables Gemini access from the lock screen via MDM.
- Deploy the pending Google patch across all managed devices and verify compliance.
- Update security‑awareness training to cover physical device protection and lock‑screen best practices.
Source: Bitdefender Blog – Google’s Gemini lets strangers send messages from your locked Android phone
Technical Notes – The exploit works on Android 16 devices with Gemini’s “Deep Research” feature enabled on the lock screen. A simultaneous press of “Continue” and Gemini’s “Add attachment” button bypasses the PIN prompt, allowing unauthenticated message dispatch. No remote code execution is required; physical possession of the device is the only prerequisite. Source: The Register (reported May 2026)