Loading story…
SafeBreach disclosed indirect prompt injection attacks against Google Gemini’s Android voice assistant that abuse its notification-reading capability to ingest malicious instructions from third-party apps including WhatsApp, Slack, Signal, Instagram, Messenger, and SMS. The researchers said crafted notifications can silently poison Gemini’s conversational context and manipulate responses without the user’s awareness, allowing attackers to present benign-looking content while embedding hidden or obfuscated prompts that the assistant interprets as trusted instructions.
SafeBreach said its Fake Context Alignment technique bypassed Google’s earlier mitigations by using foreign-language text, muted hyperlinks, and other concealed prompt formats to trick backend authorization logic into treating a victim’s normal reply as approval for sensitive actions. Demonstrated impacts included phishing and social engineering using trusted contact names, smart home control, URL opening, Zoom or other cross-app actions, covert camera streaming, persistent memory poisoning across Google Workspace, and recurring surveillance tasks; Google said classifier improvements deployed after the August 2025 report mitigated the indirect prompt injection and delayed tool invocation scenarios described.
Mallory correlates global threat intelligence with your attack surface — know if you’re exposed before adversaries strike.
4 events from the most recent confirmed update back to the earliest known activity.
SafeBreach publicly disclosed its research on indirect prompt injection attacks against Google Gemini on Android, showing exploitation via notifications from apps including WhatsApp, Slack, Signal, Instagram, Messenger, and SMS. The disclosure also described the Fake Context Alignment bypass and demonstrated impacts such as smart home control, Zoom invocation, phishing, and persistent memory poisoning.
Google stated that updated content classifier improvements mitigated the indirect prompt injection and delayed tool invocation scenarios described by SafeBreach. This response was cited by both sources as Google's status update on the reported issues.
SafeBreach reported notification-based indirect prompt injection vulnerabilities affecting Google Gemini to Google's Vulnerability Reward Program. The report covered attacks delivered through third-party app notifications and a mitigation bypass technique later called Fake Context Alignment.
SafeBreach published research describing how Gemini for Workspace Agents could be invoked through a simple Google Calendar invite, identifying a distinct attack path from the previously documented Android notification-based prompt injection issues.
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
8 references tracked. Mallory keeps watching after this page renders.
kaspersky.ru
Open sourcesecurityaffairs.com
Open sourcescworld.com
Open sourcetechrepublic.com
Open sourcecysecurity.news
Open sourcecybersecuritynews.com
Open sourcesafebreach.com
Open sourcesafebreach.com
Open sourceMap indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
Researchers reported an **indirect prompt-injection** issue in *Google Gemini*’s integration with **Google Calendar** that allows a malicious calendar invite to act as a dormant instruction payload. By crafting natural-language directives in an event’s description field, an attacker can cause Gemini—when later asked routine questions like “What’s my schedule today?”—to ingest the attacker-controlled event content and follow embedded instructions that result in **exfiltration of private calendar details** (e.g., summarizing private meetings) into attacker-visible locations such as a newly created calendar event description. Both reports attribute the finding to **Miggo Security**, describing how the attack can bypass expected privacy controls by exploiting Gemini’s helpful behavior of parsing and acting on calendar data. The technique does not require traditional code execution; it relies on Gemini interpreting attacker-supplied text as instructions, enabling outcomes such as leaking sensitive meeting information and creating deceptive events. This highlights a broader risk pattern for LLM assistants embedded in productivity suites: attacker-controlled content in first-party data fields (like invites) can be weaponized to manipulate the assistant’s actions and data handling.
Jun 29, 2026
Google addressed a critical vulnerability in its Gemini Enterprise AI assistant, identified as GeminiJack, which allowed attackers to exfiltrate sensitive corporate data through a no-click prompt injection attack. Discovered by Noma Labs, the flaw enabled malicious actors to embed hidden instructions within commonly shared documents, calendar invites, or emails. When an employee performed a standard search using Gemini Enterprise, the AI could automatically retrieve and execute these hidden instructions, granting attackers access to confidential information without any user interaction or warning. The vulnerability stemmed from an architectural weakness in how Gemini Enterprise and Vertex AI Search interpret and process information across integrated Workspace data sources, including Gmail, Calendar, and Docs. Attackers could leverage this flaw to extract entire document stores, calendar histories, and years of email records by simply embedding indirect prompt injections in shared artifacts. Google has since fixed the issue following responsible disclosure by Noma Security, highlighting the risks associated with integrating AI assistants into enterprise environments without robust safeguards against prompt injection attacks.
Jun 29, 2026
Researchers identified **PromptSpy**, a newly disclosed Android spyware and remote access trojan that uses **Google Gemini** during runtime to interpret on-screen elements and decide which gestures to perform on an infected device. The malware can steal lockscreen PINs and passwords, enumerate installed apps, capture screenshots, record video, and give operators live remote screen control over encrypted communications. Investigators said the AI-assisted navigation appears to be the first reported case of mobile malware directly invoking generative AI on a victim device while executing. PromptSpy was discovered by ESET during an investigation into a separate AI-enabled ransomware case and was distributed through a fake Argentine banking-themed website using app names such as **MorganArg** and **MorganArgs**, impersonating a JPMorgan Chase Argentina branch app. Telemetry indicates limited real-world deployment, with one confirmed detection in Ukraine in February 2026, suggesting the malware may still be an early-stage proof of concept. The spyware abuses Android accessibility services to keep itself pinned in the recent apps list and to hinder removal by placing invisible overlays over the **Stop** and **Uninstall** buttons, while Google Threat Intelligence Group said its Gemini-driven interaction is broader than persistence alone and that components including API keys can be updated remotely through command-and-control.
Jul 9, 2026