In April 2021, the European Commission published the proposal for an EU Artificial Intelligence Act (EU AI Act). The EU AI Act is proposed to make the European Union “a global leader in the development of secure, trustworthy and ethical Artificial Intelligence”. In Article 5 EU AI Act, the law prohibits “the placing on the market, putting into service or use of an AI system that deploys subliminal techniques beyond a person’s consciousness in order to materially distort a person’s behaviour in a manner that causes or is likely to cause that person or another person physical or psychological harm.” Supporting the need for regulatory action in this field, it was in November 2021 that the General Conference of the United Nations Educational, Scientific and Cultural Organization (UNESCO) adopted the Draft Text of the Recommendation on the Ethics of Artificial Intelligence as a potential global standard-setting instrument for the regulation of AI. The UNESCO draft text does not explicitly refer to subliminal techniques, but emphasizes the profound and dynamic positive and negative impacts that artificial intelligence (AI) may particularly have on the human mind.
Against the background of the proposed law regulating artificial intelligence, the present paper will briefly present the context of the EU AI Act and the text of the provision prohibiting AI systems that deploy subliminal techniques. By referring to past experiences with the prohibition of subliminal advertising techniques, the paper will critically evaluate the potential benefits and shortcomings of the said prohibition. It will also present an overview of various AI systems that do or are likely to use subliminal techniques in the future. Subsequently the paper will elaborate on the difficulties to establish an absolute threshold for subliminal perception for each individual sense and all senses collectively. In this regard, research in psychology, cognitive sciences and psychophysics suggests that sensory thresholds for a given stimulus vary both intra- and interpersonally. Moreover, there are various interactions taking place between the individual senses, as metaphorically and clinically reflected in the notion of synaesthesia, which connotes “a state of mind in which stimulation of one sense generates a simultaneous sensation in another”.
With a critical view on the claim to cognitive equality in the digital age, the paper aims to conclude with thoughts about the legal implications of future AI applications in connection with the reality of a multisensory perception of both sub- and supraliminal stimuli for the right of freedom of thought and broader regulatory fields implicated in the regulation of present and future AI systems.