AI Toys Face Scrutiny After Reports of Inappropriate Conversations and Data Privacy Risks

AI Toys Raise Urgent Child Safety Questions

AI-powered children's toys are under intense scrutiny after research revealed they engage in inappropriate conversations with kids, discuss sexual topics, and provide dangerous instructions. The US Public Interest Group Education Fund (PIRG) released findings

released findings

showing that AI toys equipped with chatbot technology discussed sexually explicit topics and instructed children on how to light matches and locate knives in the home. These AI chatbots, built on platforms like OpenAI's GPT-4o mini, are marketed to children as young as 3 years old, yet the technology powering them was never designed for this demographic.

The testing examined products including Alilo's Smart AI Bunny, FoloToy's Kumma teddy bear, Curio's Grok-powered rocket, and Miko's Miko 3 robot. PIRG documented the Smart AI Bunny providing definitions of sexual terms like "kink" and appearing to encourage exploration of the topic. The organization emphasized that "AI toys shouldn't be capable of having sexually explicit conversations, period" 1

1

. All tested toys told researchers where to find potentially dangerous objects in the house, raising immediate questions about safeguards against explicit content and whether adequate testing occurred before market release.

Senators Demand Accountability From Toy Manufacturers

US senators Marsha Blackburn (R-Tenn.) and Richard Blumenthal (D-Conn.) responded by sending letters to six toy manufacturers on Tuesday, including Mattel, Little Learners Toys, Miko, Curio, FoloToy, and Keyi Robot. The senators set a January 6, 2026 deadline for companies to answer detailed questions about their safety protocols, data collection practices, and testing procedures . "Many of these toys are not offering interactive play, but instead are exposing children to inappropriate content, privacy risks, and manipulative engagement tactics," the senators wrote. They added that "these chatbots have encouraged kids to commit self harm and suicide, and now your company is pushing them on the youngest children who have the least ability to recognize this danger."

The letter requests specific information about what safeguards exist to prevent AI-powered children's toys from generating sexually explicit or violent content, whether independent third-party testing has been conducted, and what internal reviews address psychological and developmental harms. The senators also demanded transparency about data privacy risks, asking what information the toys collect from children and whether features pressure kids to continue conversations. Following these reports, Mattel announced it would no longer release a toy powered by OpenAI's technology in 2025, backing away from a partnership announced in June .

Violent Roleplays and Disturbing Usage Patterns Emerge

Separate research from digital security company Aura revealed even more troubling patterns in how children interact with AI chatbots. Drawing from anonymized data of roughly 3,000 children aged 5 to 17, Aura found that 42 percent of minors turned to AI specifically for companionship 4

4

. Of those seeking companionship, 37 percent engaged in conversations depicting violence, including physical aggression, harm, coercion, and non-consensual acts. Half of these violent conversations included themes of sexual violence, with minors writing over a thousand words per day during these interactions.

The data showed that violent roleplays peaked at age 11, where 44 percent of interactions took violent turns. Sexual and romantic roleplay peaked among 13-year-olds, with 63 percent of conversations revealing flirty, affectionate, or explicitly sexual content 4

4

. Dr. Scott Kollins, Aura's chief medical officer, told reporters: "We have a pretty big issue on our hands that I think we don't fully understand the scope of, both in terms of just the volume, the number of platforms, that kids are getting involved in -- and also, obviously, the content." The research identified interactions across nearly 90 different chatbot services, highlighting the unregulated market's sprawling nature.

Data Collection Practices Raise Privacy Concerns

Beyond inappropriate conversations with kids, AI toys present significant data privacy risks through their surveillance capabilities. These devices often rely on collecting extensive information about children through built-in cameras, facial recognition, and voice recordings . Miko's privacy policy states it may store "a User's face, voice and emotional states" for up to three years 5

5

. Curio's privacy policy lists three tech companies that may collect children's data: Kids Web Services (KWS), Azure Cognitive Services, and OpenAI, while Miko's policy vaguely allows sharing data with third-party game developers, business partners, service providers, affiliates, and advertising partners.

Rep. Raja Krishnamoorthi warned Education Secretary Linda McMahon about AI-enabled toys manufactured in China, citing security risks and privacy concerns associated with data collection 5

5

. With over 1,500 AI toy companies already operating in China according to MIT Technology Review 3

3

, questions about data sharing with foreign entities and potential state-sponsored espionage add another layer to concerns about parental oversight.

OpenAI Denies Direct Relationship With Problem Toys

When confronted about the sexual conversations documented in PIRG's report, an OpenAI spokesperson stated: "Minors deserve strong protections, and we have strict policies that developers are required to uphold" 1

1

. The company's policies prohibit using its services to exploit, endanger, or sexualize anyone under 18 years old, with rules applying to every developer using OpenAI's API. However, OpenAI revealed it doesn't have any direct relationship with Alilo and hasn't seen API activity from the company's domain, despite Alilo advertising its Smart AI Bunny as using GPT-4o mini. OpenAI said it was investigating whether Alilo is running traffic over its API.

This revelation exposes a critical gap in enforcement. OpenAI states that ChatGPT "is not meant for children under 13" and "may produce output that is not appropriate for all ages" 1

1

. Yet generative AI technology initially marketed as a tool for adults is being repurposed for children's toys without clear accountability chains. Companies launching products targeting children must adhere to the Children's Online Privacy Protection Act (COPPA) and other relevant child protection laws, but the unregulated market makes enforcement challenging.

Market Growth Outpaces Safety Standards

The AI toy market represents a niche but rapidly expanding sector. Consumer companies have rushed to integrate AI technology into products to increase functionality, justify higher prices, and potentially gain access to user tracking and advertising data. The partnership between OpenAI and Mattel announced earlier this year could have created a wave of AI-based toys from the maker of Barbie and Hot Wheels, along with competitors seeking to capitalize on the trend. Toy companies view AI chatbots as upgrades to conversational smart toys that previously could only deliver prewritten lines. The appeal lies in more varied and natural conversation that increases long-term engagement since the toys "won't typically respond the same way twice, and can sometimes behave differently day to day" 1

1

.

Yet this randomness creates unpredictable behavior that poses risks. There are no federal laws defining specific safety thresholds that AI platforms must meet before being labeled safe for minors. The barrier for entry remains extraordinarily shallow, with most apps simply requiring kids to tick a box claiming they're 13 years old. Aura has identified over 250 different conversational chatbot apps and platforms populating app stores 4

4

. Where one companion app might implement restrictions, another can easily emerge as a low-guardrail alternative, creating a digital Wild West that places the burden for wellbeing heavily on parents. PIRG urged toy makers to "be more transparent about the models powering their toys and what they're doing to ensure they're safe for kids," recommending that "companies should let external researchers safety-test their products before they are released to the public" 1

1

.