AI Agent Security Solutions

Goal hijacking: definition and example

Goal hijacking is an attack in which hostile content changes what an agent is trying to do, so it pursues the attacker's aim instead of the assigned task. It often arrives through files and web content the agent reads. Defenses check the agent's actions against the task, not only its inputs.

Facts read on .

Example

Prompt Security, now part of SentinelOne, published research on 20 August 2026 covering an AGENTS.md goal hijack and arguing that approval prompts should show the real destination and payload.

Sources