New AI Workflow Identity Hijacking Attack Lets Hackers Exfiltrate Sensitive Data
A new attack tricks AI-powered business tools into giving attackers access to private company files and messages. Attackers send fake requests to these automated workflows to steal sensitive data.
- Report priority
- Medium
How it works
- Attackers send fake requests to AI-powered business tools that handle company files and messages.
- These tools trust the requests and share private data with the attacker instead of the real user.
- The flaw lets attackers bypass normal security checks and grab data they should not see.
What to do
If your company uses AI tools to automate tasks like moving files, sorting messages, or processing data, check if those tools are part of a business workflow (like Slack, Teams, or custom scripts). If you manage or own such tools, review their security settings and recent updates for fixes. If you are a user, ask your IT team if these tools were updated or patched recently.
Contact your AI tool provider or IT team to confirm if a security update is available. Ask if your company's workflows were checked for this issue. If no update is available, restrict access to sensitive data in these tools until a fix is released. Check the provider's normal update channel or support portal for details on the fix.
Technical details
Security researchers have recently disclosed a new enterprise AI attack technique known as Workflow Identity Hijacking. This method enables external attackers to exfiltrate sensitive corporate information by submitting seemingly harmless requests to AI-powered automations.