Security, Affective Computing, and the Protection of the Free Mind

Emotional AI systems decode human emotions through facial micro-expressions, vocal patterns, and physiological data. While the EU Artificial Intelligence Act prohibits emotion recognition in workplaces and educational institutions, broad security exemptions permit deployment in military, defense, national security, and law enforcement contexts. This article demonstrates how these exemptions create normalization pathways through which surveillance technologies migrate from exceptional security applications into routine governance. Moving beyond commonly examined privacy and data-protection frameworks, we advance freedom of thought under Article 9 ECHR as the principled basis for an absolute prohibition. Emotional AI technologies extract involuntary, pre-cognitive affective states that shape conscious belief formation, thereby challenging the binary distinction between forum internum and forum externum protection. Building on evolving ECtHR jurisprudence on non-disclosure rights, we propose ‘affective integrity’: the right to experience emotions free from technological surveillance and inference, requiring absolute protection immune from security justifications and regulatory exemptions.