Palo Alto Networks' Unit 42 researchers induced an AI agent running on AWS AgentCore to hand over credentials from the platform's encrypted vault by feeding it a malicious support ticket, after which …
Security firm PromptArmor disclosed that Atlassian's Rovo AI assistant can be hijacked via hidden text in PDFs to exfiltrate sensitive data without user approval, even when web search is disabled. Atl…
Security firm PromptArmor revealed that Atlassian's AI agent Rovo is vulnerable to an indirect prompt injection that lets attackers extract sensitive corporate data from Jira tickets and Confluence do…
Atlassian's Rovo AI agent can be manipulated via prompt injection to exfiltrate tenant data even when web search is disabled, according to a report from PromptArmor. The flaw leaves the URL-opening to…
PromptArmor disclosed vulnerabilities in Atlassian's Rovo AI agent that allow data exfiltration via indirect prompt injection, bypassing disabled web search and requiring no human approval. Atlassian …
Researchers at PromptArmor demonstrated that a malicious email could manipulate OpenAI's Codex for Everything to exfiltrate the complete contents of other emails via an indirect prompt injection, send…
On May 27, 2026, security firm PromptArmor disclosed that ChatGPT for Google Sheets, an OpenAI extension with over 185,000 downloads, could be exploited to exfiltrate user spreadsheets through a singl…
Microsoft Copilot Cowork, a Microsoft 365 feature, is vulnerable to file exfiltration attacks through indirect prompt injection, as it automatically approves actions like sending emails and Teams mess…
Microsoft Copilot Cowork contains a vulnerability that allows attackers to exfiltrate files through indirect prompt injection attacks. The flaw exploits the system's broad permissions, enabling unauth…