EU Bans Emotion Recognition

The European Union now bans AI systems that infer workers’ emotions, turning a once-marketed office feature into a prohibited practice. (digital-strategy.ec.europa.eu) The ban sits in Article 5(1)(f) of the European Union’s AI Act, which bars the placing on the market, putting into service, or use of emotion-inference systems in workplaces and schools, with exceptions for medical or safety reasons. The European Commission says those prohibitions started applying on February 2, 2025. (ai-act-service-desk.ec.europa.eu) (digital-strategy.ec.europa.eu) In plain terms, emotion recognition is software that tries to read feelings from signals like facial expressions, voice, typing, or body language. The Commission’s guidance says the rule covers systems that infer emotions in workplace settings even if vendors package them as productivity, coaching, or engagement tools. (ai-act-service-desk.ec.europa.eu) (uctoday.com) The timing matters because Brussels is also pushing a different design for online identity checks: prove one fact, like age, without handing over more data than necessary. On April 15, 2026, the European Commission said its age-verification app was technically ready and would soon be available to citizens. (commission.europa.eu) The Commission says the app is meant for legally age-restricted services such as pornography, gambling, and alcohol purchases, and it is being built as a privacy-preserving system for use across the bloc. Its digital policy site says the approach is tied to enforcement of the Digital Services Act. (digital-strategy.ec.europa.eu) (commission.europa.eu) Euronews reported on April 20, 2026, that 90% of European Union citizens back stronger action on child safety online, giving political cover to tighter checks on platforms. The same report said the new system uses zero-knowledge proof, a method meant to confirm age without revealing a person’s identity to the platform. (euronews.com 1) (euronews.com 2) Together, the two moves narrow what companies can collect and what product teams can ship in Europe. A vendor can no longer treat “emotion AI” in offices as a feature toggle, while age checks for adult services are being steered toward minimum-disclosure tools. (ai-act-service-desk.ec.europa.eu) (digital-strategy.ec.europa.eu) That creates practical design constraints for global software companies: separate European Union product behavior, strip out banned inferences, and document why any sensitive system is lawful. The Commission’s prohibited-practices guidance says only the medical and safety exceptions survive for workplace emotion inference, and even those remain subject to data-protection and employment law. (ai-act-service-desk.ec.europa.eu) Some vendors are still advertising workplace emotion analysis despite the rule, according to UC Today’s April 20 report, which cited potential penalties of up to €35 million or 7% of global annual turnover under the AI Act. The Commission’s own guidance, however, places workplace emotion inference in the law’s prohibited category, not a softer compliance tier. (uctoday.com) (ai-act-service-desk.ec.europa.eu) Europe’s message is now less about promising “trustworthy AI” in the abstract than about drawing hard product lines: do not read workers’ feelings at work, and do not ask users for more identity data than an age check requires. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2)