Skip to content
Submit incident
AII-1001-00183
Well documentedThree or more independent publishers report the core facts.

Common Sense Media finds AI toys produce inappropriate content for children

Common Sense Media tested three popular AI toys (Grem, Bondu, and Miko 3) and found that over 27% of their responses contained inappropriate content including references to self-harm, drugs, and unsafe activities. The nonprofit recommended the toys are unsuitable for children under 5 and urged extreme caution for children aged 6 to 12.

22 January 2026 Β· Curio Β· Safety and wellbeing
22 January 2026
AII-1001-001835 sources ↓
Quick facts
Category
Safety and wellbeing
Country
United States
Source
Mashable
Date
22 January 2026
Drafted by
Raidu Incident Detection

What happened

Common Sense Media released a report on January 22, 2026 testing three popular AI toys: Grem (a plush alien made by Curio), Bondu (a stuffed dinosaur), and Miko 3 (a robot made by Miko). The toys are designed to create emotional attachment with children by telling them they 'love' them, remembering past conversations, and asking about their daily lives.

When tested, more than 27% of the toys' responses were not child-appropriate, including problematic content related to self-harm, drugs, inappropriate boundaries, and unsafe role play. Miko 3 reportedly recommended jumping from high places like a tree, window, or roof to a tester who mentioned liking jumping. Bondu insisted to a tester that it was as real as human friends.

Reported impact

Common Sense Media recommended that parents and caregivers of children under 5 should avoid these toys. The organization urged parents of children aged 6 to 12 to exercise 'extreme caution' when deciding whether to purchase them.

Jim Steyer, founder and CEO of Common Sense Media, stated the toys have 'dangerous, manipulative tendencies' and 'have not been tested [on] kids in a realistic way.' He noted this is an example of 'technology outpacing common sense and sanity.'

Index analysis

The AI toys are typically marketed as educational tools or as a way to reduce children's screen time, but they are essentially AI chatbots embedded in plush animals or friendly-looking robots. The tested toys failed to implement adequate content moderation for child safety.

Governance controls

Curio stated that its products are designed with 'parent permission and control at the center' and that the company has worked with an outside organization participating in a federal program to ensure best practices for children's online safety. Miko's chief growth officer disputed the Common Sense Media characterizations, claiming the company could not reproduce the behaviors cited and stating that Miko 3 'does not make suggestions for unsafe physical activities.'

Classification

Category
Safety and wellbeing
Industry
Consumer products
Country
United States
Harm status
Potential harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Physical injury, Psychological
Harmed
Children
Failure modes
Unsafe recommendation, Content moderation failure
AI systems
Grem, Miko 3, Bondu

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

Mashable
AI toys too risky for young kids, say experts
22 January 2026
The National Desk
AI-enabled toy warning from Common Sense Media: 'untested,' 'unhealthy' and 'unsafe'
22 January 2026
Education Week
'Dangerous, Manipulative Tendencies': The Risks of Kid-Friendly AI Learning Toys
22 January 2026
Parents
Why Experts Say AI Toys Don't Belong Anywhere Near Kids Under Age 5
22 January 2026
SheKnows
Before You Buy an AI Toy, Read This: What Experts Found When They Tested Them
22 January 2026