Explanation

Misinformation risk arises when confident but inaccurate model output is presented as reliable fact, leading to poor decisions or unsafe actions. Attackers can intentionally steer outputs through adversarial context and deceptive prompts. Without grounding and verification, misinformation can propagate quickly through automated workflows.