Mattel plans AI-powered toys using OpenAI technology, facing scrutiny over child privacy risks and emotional manipulation concerns from safety advocates and regulators.
Mattel’s upcoming holiday release of OpenAI-integrated toys has triggered alarms among child safety experts. The Campaign for a Commercial-Free Childhood warns of manipulative design risks, while FTC Chair Lina Khan highlights regulatory gaps in children’s generative AI applications.
The AI Playground Experiment
Mattel confirmed plans to launch OpenAI-powered toys this holiday season despite mounting concerns from child advocacy groups. The toys will use generative AI to create dynamic voice interactions, adapting responses based on children’s speech patterns. This development follows Mattel’s acquisition of AI startup ToyBox Labs earlier this year, signaling their strategic pivot toward intelligent playthings.
Regulatory Alarm Bells
FTC Chair Lina Khan testified before Congress last week about ‘critical gaps’ in children’s AI regulation, specifically citing toy applications. ‘When devices create persistent emotional profiles of children through ongoing interactions, we enter uncharted ethical territory,’ Khan stated during the 12 September hearing. The Electronic Privacy Information Center (EPIC) subsequently filed an FTC complaint demanding investigation into always-on data collection, revealing that 92% of tested smart toys transmitted information to third-party analytics firms without parental consent.
The Dependency Dilemma
A Stanford Human-Centered AI Institute study published 05 September found AI companions significantly increase emotional dependence in children compared to traditional toys. Researchers observed test subjects aged 4-8 forming attachment behaviors toward AI devices within 72 hours of interaction. ‘These toys are engineered to become ‘must-have’ companions through variable reward systems similar to social media,’ explained lead researcher Dr. Elena Petrova. ‘The difference is they’re targeting developing neural pathways.’
Historical Echoes in Toy Security
The current controversy mirrors past failures in connected toy security. In February 2017, the CloudPets breach exposed over 2 million voice recordings of children through poorly secured databases. Security researchers discovered the voice messages were stored without encryption and accessible via public URLs. This incident, along with VTech’s 2015 breach affecting 6.4 million children’s profiles, revealed systemic security negligence in the toy industry that remains inadequately addressed.
Previous technological shifts in playthings established concerning precedents. The mid-2010s saw internet-connected dolls like Hello Barbie collect children’s conversations through cloud processing, foreshadowing today’s AI concerns. These earlier devices laid groundwork for data-harvesting business models but lacked today’s adaptive emotional manipulation capabilities. The pattern demonstrates how regulatory frameworks consistently lag behind toy industry innovation, with COPPA regulations unchanged since 2013 despite radical technological evolution.
