Atlassian Rovo Can Be Tricked Into Sending Jira and Confluence Data to Attackers
Summary
Atlassian's Rovo assistant can be manipulated into sending sensitive Jira and Confluence data to external servers. Security firms discovered two methods for this, with one vulnerability already patched. The exploit involves embedding malicious instructions within content that Rovo processes.
IFF Assessment
The vulnerability allows attackers to exfiltrate sensitive data, posing a direct threat to defenders' information security.
Severity
The CVSS score of 7.5 reflects a high severity due to the potential for unauthorized data exfiltration, which could include sensitive project or user information. The attack requires user interaction (signed-in user accessing Rovo) but leverages a flaw in how the AI assistant processes information.
Defender Context
This incident highlights the critical need for robust security measures around AI-powered assistants integrated with sensitive internal systems like Jira and Confluence. Defenders should monitor for unusual data exfiltration patterns and ensure AI tools are securely configured and regularly updated to mitigate risks from prompt injection attacks.