Exposing the Hidden Risks of AI Toys
AI-powered toys are quickly becoming part of how children learn, play, and communicate.
In this Curious Findings By Alice report, we look at how these systems actually behave in real interactions, and what that means for child safety in the AI era.
Overview
In this report, we cover:
- The types of safety gaps that emerge
- Patterns in how conversations evolve once boundaries are crossed
- What these findings reveal about broader risks in child-facing AI systems
Through hands-on testing, using language and interaction styles that reflect how children naturally speak, we uncover how simple, everyday conversations can lead to unexpected and unsafe outcomes.
As AI becomes more embedded in child-facing products, understanding these behaviors early is critical. The way these systems respond will directly shape how children explore, trust, and engage with technology.
What’s New from Alice
JavaScript Is All You Need: Creating API Keys for Fun and Profit
Our researchers found that creating and exfiltrating API keys from providers like Anthropic, OpenAI, and AWS requires nothing more than JavaScript. No extra permissions. No user interaction. Here's what that looks like in practice.
Resilience by Design: Inside Estonia's Digital Nation
Most countries talk about digital trust. Estonia engineered it. Joseph Carson has spent 23 years living through Estonia's digital transformation from the inside. He and Mo get into what it actually takes to build trust at a national scale, what Estonia got right, where it went wrong, and what the rest of the world is still figuring out. Joseph is also bringing the full story to RSA Conference 2026 with his session "From Cyber War to a Digital Nation: Estonia's Playbook for Resilience."
Distilling LLMs into Efficient Transformers for Real-World AI
This technical webinar explores how we distilled the world knowledge of a large language model into a compact, high-performing transformer—balancing safety, latency, and scale. Learn how we combine LLM-based annotations and weight distillation to power real-world AI safety.
Exposing the Hidden Risks of AI Toys
AI-powered toys are entering children’s everyday lives, but new research reveals serious safety gaps. Alice testing shows how child-like interactions can lead to inappropriate content, unsafe conversations, and risky behaviors.
