New research program to coordinate studies on AI’s cognitive effects on children — filling critical knowledge gaps as kids spend hours daily with general AI models and chatbots
Worldwide, concern is growing about the early effects AI has shown to have on children. Children now send longer, more emotionally complex messages to AI chatbots than to their own friends – averaging 163 words to AI versus just 12 words to peers[i] – yet researchers still don’t know enough about how these intimate digital relationships affect developing brains.
Today, the Childhood and AI Lab announced a Research Fellowship program to answer that question — the first dedicated network for researchers studying AI’s cognitive and developmental effects on children.
While governments debate high-level frameworks, a critical gap remains: the foundational research needed to turn principles into practice. Schools are creating AI policies without knowing if AI tutors help or harm learning. Tech companies face pressure to prove their products are safe for children but lack developmental benchmarks to test against. And parents are making daily decisions about AI use with almost no scientific evidence to guide them.
Unlike existing AI safety research, which focuses primarily on preventing harmful content, this new Fellowship program examines something harder to measure but potentially more consequential: how AI changes the way children think, learn, and relate as they grow.
John Zoltner, founder of the Childhood and AI Lab and CEO of AIChildSafety.org said: “When a child asks AI to think for them 20 times a day, something changes in that brain. We need to know what. Every AI conversation shapes how children learn to think but right now, we’re guessing at the effects. We created the Fellowship to end the guessing. We’re searching for answers, not more talk or speculation.
“We have decades of research on screen time, social media, and video games. We need that same rigor for AI — and we need it urgently. Governments, tech companies, and regulators will need real-time evidence to make design and policy decisions that truly keep children safe. These tools are already in classrooms and living rooms around the world.”
The Fellowship
The initiative addresses what researchers and officials call a “policy-evidence gap” — regulators and companies are making decisions about children’s AI use without foundational research on developmental impacts.
The Fellowship aims to build a new generation of researchers who understand both child development and AI technology – experts who can answer questions no single discipline can tackle alone. How does daily AI use rewire attention spans? What happens to creativity when children rely on algorithms to generate ideas? Does AI tutoring strengthen or undermine learning?
Fellows will investigate these questions across neuroscience, education, computer science, and ethics, collaborating through cross-discipline collaboration, joint research projects and regular symposia. The goal isn’t just more research papers – it’s actionable guidance that parents can use at the dinner table, teachers can apply in classrooms, and policymakers can write into law.
The Childhood and AI Lab is recruiting a wide range of fellows whose research will lead to new capabilities, such as:
- Measurable benchmarks for cognitive safety (e.g., testing whether AI tutors that give instant answers reduce problem-solving persistence)
- Assessment tools that schools and parents can use to evaluate AI products
- Age-appropriate design standards based on developmental psychology, not just content filters.
The Lab has already secured partnerships with the University of Western Sydney in Australia and Middlesex University in the United Kingdom.
The Childhood and AI Lab will begin accepting applications from Wednesday 5 November 2025. Information about the Fellowship and the online application can be found at http://www.childhoodandailab.org
[i] Source: Kids and AI: Aura Finds Kids Turning to AI in Disturbing Ways
