AI increasingly selects, generates, and frames information before it reaches a person. Cognitive security asks how people can preserve informed, autonomous judgment under that mediation.
A system can be technically secure while still shaping perception in ways that are difficult to recognize. People need ways to detect manipulation, understand why information is presented, calibrate trust, and compare independent evidence.
This is a research direction we are developing. We are defining the threats, mechanisms, and methods needed to make these risks visible and preserve human agency.