EU Proposes Ban on Emotional AI Chatbots for Children

The European Union has unveiled sweeping restrictions on how artificial intelligence chatbots can interact with children, targeting what regulators describe as emerging risks of emotional dependency and developmental harm.

Under the proposed EU Kids Act introduced Thursday, chatbots would be prohibited from simulating emotions or interpersonal relationships when interacting with anyone under 18. Children under 13 would be banned from accessing chatbots entirely without guardian supervision, according to draft legislation reviewed by WIRED, which first reported the details.

Strict Memory and Design Requirements

The proposed rules extend beyond age verification. Chatbots operating in the EU would be barred from retaining memory between separate conversations with minors—a measure designed to prevent systems from accumulating sensitive data over time and reinforcing potentially harmful interaction patterns.

Digital platforms would also face restrictions on how they present AI chatbots to young users. The legislation would prohibit automatically activating chatbots or encouraging their use, including practices like pinning AI companions at the top of a minor's contact list—a feature Snapchat had previously implemented.

Companies that fail to build child-safety protections into their products could face fines reaching 6% of total worldwide turnover.

Why It Matters

This represents the first major regulatory framework specifically targeting AI chatbot interactions with minors. As conversational AI becomes embedded in social platforms and educational tools, the EU is establishing a precedent that could influence global standards for child safety in AI systems. The financial penalties are substantial enough to force compliance from major technology companies operating in European markets.

Concerns About Development and Mental Health

European Commission documents cite worries that chatbot dependency may negatively impact child development and relationships with other humans. Regulators also expressed concern about chatbots providing dangerous advice on mental health and personal development that could encourage risky behavior.

President Ursula von der Leyen stated at Thursday's press conference that "platforms have been playing fast and loose with children's mental health." Survey data from Ipsos BVA indicates that roughly half of Europeans aged 11 to 25 have discussed intimate or personal matters with AI chatbots across France, Germany, Sweden, and Ireland.

The Commission also pointed to research from the Organisation for Economic Co-operation and Development spanning 91 countries, which found that children who use AI chatbots for schoolwork score lower on tests than those who don't.

Scope and Future Expansion

The proposed restrictions apply specifically to AI companions and general conversational chatbots that the Commission considers high-risk for minors. Simeon de Brouwer, a policy adviser at European Digital Rights, questioned why protections are limited to young people, asking whether all users deserve protection from addictive AI design and emotional dependencies.

The EU indicated its Digital Fairness Act, expected later this year, will address addictive design patterns more broadly across user demographics.

The legislation also addresses deepfake concerns, particularly AI-generated sexualized images of minors—an issue von der Leyen referenced in Wednesday's State of the Union speech. While the EU AI Act has required labeling of synthetic content since August, current rules only cover deepfakes created for commercial purposes.

WIRED first reported the details of the proposed EU Kids Act.