Applicable to: AI providers and deployers.
Date of application: 2 February 2025.
Objective: to prohibit the use of AI systems that manipulate individuals through subliminal or deceptive techniques causing them significant harm.
Regulation (EU) No 1689/2024 on Artificial Intelligence (AI Act) of 13 June 2024 introduced a series of prohibitions within the European Union on practices involving the use of AI systems that pose an ‘unacceptable’ risk (the highest level assigned to certain AI systems in the risk-based approach adopted by the AI Act).
In this article, we will analyse the practice covered by Article 5(1)(a) of the AI Act, under which the following is prohibited:
the placing on the market, putting into service or use of an AI system that deploys subliminal techniques beyond a person’s consciousness or purposefully manipulative or deceptive techniques with the objective, or the effect of materially distorting the behaviour of a person or group of persons, by appreciably impairing their ability to make an informed decision, thereby causing them to take a decision they would not have otherwise taken in a manner that causes or is reasonably likely to cause to that person, another person or a group of persons significant harm.
The elements that characterise the prohibited practice are:
1) the fact that the AI system is placed on the market, put into service or used within the European Union
2) that it makes use of subliminal techniques that operate without a person being aware of them, or techniques that are deliberately manipulative or deceptive
3) that the techniques are intended to, or have the effect of, materially distorting the behaviour of a person or a group of persons, significantly impairing their ability to make an informed decision, and leading them to make a decision they would not otherwise have made
4) that the decision taken causes, or is reasonably likely to cause, significant harm to that person, another person or a group of persons.
As specified in the Guidelines approved by the European Commission on 29 July 2025, ‘subliminal’ techniques include visual and audio subliminal messages (which, respectively, appear too quickly or are transmitted at a too low volume to be registered by the brain), as well as images embedded in other visual content.
The concept of ‘deliberately manipulative or deceptive’ techniques, on the other hand, includes those that undermine decision-making autonomy by presenting false or misleading information (for example, an AI chatbot that uses a synthetic voice to pretend to be a friend or relative of a person, carrying out a scam).
The ban applies both where the system is used with the intention of distorting users’ behaviour, and where such distortion is merely a consequence of the use of subliminal techniques that are deliberately manipulative or deceptive.
To clarify the above, we can take the example of a photographer who uses an AI system to alter photographs of people in order to accentuate their emotions, expressions or characteristics with the aim of influencing purchases, political choices or other behaviours.
In this case, the activity may fall within the scope of the prohibition if the individual whose choices have been influenced by the alterations made to the photographs has suffered significant harm.
The concept of ‘significant harm’, which is necessary for the prohibition to apply, is not defined within the AI Act. The Guidelines have, however, specified that it must result in a significant negative impact on physical or psychological health, or on the financial and economic interests, of a person or a group of people.
Articoli Correlati