In the name of security (against science): AI Act and emotion recognition

The regulation of emotion recognition in the EU AI Act shows a dangerous reversal of the precautionary principle. Despite the premise in recital 44 – “there are serious concerns about the scientific basis of AI systems aiming to identify or infer emotions […]” - emotion recognition is prohibited only in very delimited cases. This policy choice is one of many indicators of the gradual weakening, even in the European Union, of the pillars of a democratic society.