Prohibited Practice

The deployment of an AI system that deploys subliminal techniques beyond a person's consciousness or purposefully manipulative or deceptive techniques to materially distort behavior.

• Mechanism: Exploits cognitive biases or sensory stimuli a person cannot perceive.
• Effect: Causes or is likely to cause physical or psychological harm.
• Example: Inaudible voice commands embedded in audio to influence purchasing decisions without the user's knowledge.

The use of an AI system that exploits vulnerabilities of a specific group of persons due to their age, disability, or a specific social or economic situation.

• Target: The elderly, children, or persons with mental or physical disabilities.
• Objective: To materially distort the behavior of a member of that group.
• Outcome: Causing or likely to cause that person or another person physical or psychological harm.

The placing on the market, putting into service, or use of AI systems for the evaluation or classification of natural persons over a certain period of time based on their social behavior or known, inferred, or predicted personal characteristics.

• Scope: Covers both public authorities and private actors.
• Condition: The social score leads to detrimental or unfavorable treatment that is either unrelated to the context where the data was generated or is unjustified or disproportionate.
• Example: A government system that assigns a 'trustworthiness' score based on online activity, leading to denial of travel visas.

The use of AI systems for real-time remote biometric identification in publicly accessible spaces for the purpose of law enforcement.

• Technology: Typically involves facial recognition against a watchlist.
• Prohibition Scope: Covers the 'real-time' capture and matching of biometric data.
• Strict Exceptions: Narrowly defined exceptions exist for targeted searches for specific crime victims, prevention of imminent terrorist threats, or localization of suspects for specific serious crimes, but require prior judicial or independent administrative authorization.

The use of AI systems to make risk assessments of natural persons to predict the likelihood of them committing a future criminal offense.

• Basis: Prohibited when based solely on the profiling of a natural person or on assessing their personality traits and characteristics.
• Distinction: This does not ban AI systems used to support human assessment of a person's involvement in a criminal activity that is already based on objective, verifiable facts directly linked to a crime.

The placing on the market, putting into service, or use of AI systems to infer the emotions of a natural person in the areas of workplace and education institutions.

• Exception: The prohibition does not apply where the use is for medical or safety reasons, such as monitoring a pilot's fatigue levels.
• Rationale: The scientific basis for inferring complex emotional states from biometric data is contested, and the practice poses high risks of discrimination and psychological pressure in power-imbalanced environments.

Deploying an AI system that deploys subliminal techniques beyond a person's consciousness to materially distort their behavior in a manner that causes or is likely to cause physical or psychological harm. This targets systems using imperceptible audio or visual cues to bypass rational decision-making.

• Mechanism: Flashing images or masked audio signals.
• Key Distinction: The manipulation must be subliminal and harmful.

Using an AI system that exploits vulnerabilities of a specific group of persons due to their age or disability to materially distort their behavior in a harmful manner. This targets predatory systems designed to manipulate children or individuals with cognitive impairments.

• Target Groups: Minors, persons with mental or physical disabilities.
• Example: A toy chatbot designed to encourage dangerous dares.

The use of AI systems by public authorities for the evaluation or classification of the trustworthiness of natural persons over a certain period based on their social behavior or known personality characteristics, leading to detrimental treatment. This bans generalized, government-led social credit systems.

• Prohibited Context: Detrimental treatment unrelated to the original data context.
• Scope: Applies to both public and private actors if the scoring leads to disproportionate social consequences.

The use of real-time remote biometric identification systems in publicly accessible spaces for the purpose of law enforcement, unless strictly necessary for specific, exhaustive exceptions. This prohibits mass surveillance via live facial recognition in public squares.

• Technology: Live facial recognition, gait analysis.
• Exceptions: Targeted searches for victims of kidnapping, prevention of imminent terrorist threats, or localization of serious criminal suspects.

AI systems used to assess the risk of a natural person committing a criminal offense based solely on profiling or personality traits. This bans pre-crime assessments derived from behavioral patterns rather than objective, verifiable facts directly linked to criminal activity.

• Prohibited Basis: Profiling, past behavior, personality traits.
• Allowed Basis: Objective and verifiable facts directly linked to a specific criminal activity.

The use of AI systems to infer the emotions of natural persons in the areas of workplace and education institutions, except where strictly justified for medical or safety reasons. This prohibits the continuous monitoring of employee or student affective states.

• Prohibited Context: Employee performance reviews, student engagement monitoring.
• Exception: Therapeutic medical devices or safety-critical operator fatigue detection.