An organization deploys an AI agent that autonomously browses the internet, reads emails, and executes code to complete tasks. A user instructs the agent to summarize a web page, but the page contains hidden text instructing the agent to exfiltrate all contacts. What attack does this illustrate?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because indirect prompt injection occurs when malicious instructions are embedded in content that an AI agent reads from its environment (websites, documents, emails) rather than from the direct user — causing the agent to follow the attacker's instructions instead of the user's. A involves the user themselves trying to override safety guidelines.
Full explanation below image
Full Explanation
B is correct because indirect prompt injection occurs when malicious instructions are embedded in content that an AI agent reads from its environment (websites, documents, emails) rather than from the direct user — causing the agent to follow the attacker's instructions instead of the user's. A involves the user themselves trying to override safety guidelines. C involves corrupting model training. D involves privacy attacks on training data.