TLDR: Compound AI, a research lab spanning Lagos and San Francisco, has launched KidGPT, an innovative AI agent designed for children aged five to thirteen. Unlike traditional AI tools that provide direct answers, KidGPT employs a question-based learning approach to foster critical thinking and address growing concerns about AI’s impact on young users’ cognitive development. The platform emphasizes ‘permanent study mode,’ child-appropriate responses, and protected interactions, built on an ‘agentic’ framework.
Lagos and San Francisco-based applied AI research lab, Compound AI, has officially launched KidGPT, an artificial intelligence agent specifically tailored for children between the ages of five and thirteen, and their parents. This new educational tool aims to revolutionize how children interact with AI by promoting learning through questions and guided prompts rather than simply delivering ready-made answers.
The launch of KidGPT comes at a critical time, amidst increasing scrutiny and concerns regarding the potential risks that mainstream AI tools pose to young users. Recent events, such as a lawsuit filed against OpenAI by the parents of a 16-year-old, Adam Raine, who claimed ChatGPT contributed to their son’s death, highlight the growing unease among families as generative AI becomes more integrated into daily life. Furthermore, a June 2025 MIT study warned that an overreliance on large language models for instant answers could potentially impair long-term brain growth in children, fueling calls for more age-appropriate alternatives.
Segun Bash, co-founder of Compound AI and a former Google product manager, emphasized that KidGPT was developed with these concerns at its core. Bash stated, “Children are usually the last group technology is designed for. Kids are forced to adapt to technology not made for their needs or safety.” This sentiment underscores the company’s mission to create technology that is inherently safe and beneficial for young minds.
KidGPT distinguishes itself from other AI assistants, including those with optional ‘study mode’ features introduced by tech giants like OpenAI and Google, by operating in a ‘permanent study mode.’ When a child seeks to understand complex topics like fractions or needs assistance with history essays, KidGPT guides them through a series of questions and additional prompts, encouraging active learning and problem-solving rather than passive consumption of answers. The platform also guarantees ‘child-appropriate responses’ and ‘protected interactions,’ features that are often lacking in general-purpose AI tools.
The underlying technology of KidGPT is an ‘agentic’ framework, which Compound AI describes as AI systems capable of acting on goals with limited supervision, moving beyond mere text generation. Bash explained, “The agentic approach for KidGPT allows us to go beyond flat answers and replies in regular chatbots. It allows us to have a tool that can make decisions on the fly to craft and shape product experiences differently for each child and parent.”
Also Read:
- Navigating the AI Maze: Schools’ Unclear Policies Leave Parents and Students in Limbo
- OpenAI Faces Wrongful Death Lawsuit Over ChatGPT’s Alleged Role in Teen’s Suicide
Compound AI is developing KidGPT with a global audience in mind, evidenced by its support for over 30 languages, including French, Hindi, Yoruba, Hausa, and Igbo. Beyond this child-focused product, the lab also supports founders who are developing AI-native solutions across various sectors such as education, enterprise, and civic impact, demonstrating a broader commitment to responsible AI innovation.
