Sharing IP/PI/confidential information with user risk for AI
Description
AI agents with unrestricted access to resources or databases or tools could potentially store and share PI/IP/confidential information with system users when performing their actions.
Why is sharing ip/pi/confidential information with user a concern for foundation models?
AI agents may share privileged information to users. The act of sharing the information may result in harm for the model owner, user, or others. The harm can vary based on the type and details of the information shared. Without adequate oversight, these privacy incidents might overwhelm company resources.
Parent topic: AI risk atlas
We provide examples covered by the press to help explain many of the foundation models' risks. Many of these events covered by the press are either still evolving or have been resolved, and referencing them can help the reader understand the potential risks and work toward mitigations. Highlighting these examples are for illustrative purposes only.