Teens, AI, and the conversations we are not having

Laura Zimmermann, Ph.D., is the Director of Youth AI Research at Common Sense Media, where she leads research within the newly launched Youth AI Safety Institute — the first independent safety lab dedicated to testing how AI products affect children. She earned her Ph.D. in Psychology at Georgetown University with a concentration in Human Development and Public Policy, and was an Institute of Education Sciences Postdoctoral Research Fellow in the School of Education.

For over a decade, Laura has studied how children learn from screens — starting in infancy and early childhood and working her way up to tweens and teens. Her graduate research at Georgetown, in collaboration with Binghamton University, tested whether the well-documented "transfer deficit" (young children learn less from a 2D screen than from a live person) also applied to the then-brand-new iPad. It did. Her landmark paper, The Ghost in the Touchscreen: Social Scaffolds Promote Learning by Toddlers (Child Development, 2017), showed what closes that gap: a responsive human sitting next to the child. She is also a co-author of Developer Meets Developmentalist: Improving Industry–Research Partnerships in Children's Technology (Journal of Children and Media, 2018).

Across her career she has led research for PBS KIDS, Sesame Workshop, New America, the Aspen Institute, the Bay Area Discovery Museum, Google, and MIND Education. She credits two mentors in particular: Lisa Guernsey at New America, and Donna Mitroff — the educational consultant behind Mister Rogers' Neighborhood. At Common Sense Media she is a lead author of Teens in the AI Era: Schoolwork and Skills That Matter, a nationally representative survey of over 1,000 US teens aged 13–17, fielded with NORC at the University of Chicago.

In this episode, Laura talks with Svenia Busson about:

• The transfer deficit — why toddlers learn less from a touchscreen than from a person, and the one thing that closes the gap
• Why 70% of teens now use AI for schoolwork, but only a minority have ever discussed with a teacher what AI actually is
• The 63% who get answers from AI — and the 77% of that same group using it for brainstorming, feedback and checking work instead
• The nearly 90% of US teachers who have received no AI professional development, and what that vacuum produces
• Why 44% of teens have had an AI tool blocked at school — and 59% of them simply switched to a personal device
• The 38% who say AI makes them generate fewer of their own ideas, and keep using it anyway — grind culture as a hypothesis
• AI companions: 72% of teens have used one, and why Common Sense Media's position is that under-18s shouldn't use them at all
• The sycophancy problem — why "practising" social skills with a character that always agrees with you isn't practice
• The equity question nobody has data on yet: does AI narrow or widen the gap between well-resourced and under-resourced students
• The developer/developmentalist gap — what it would take for child development evidence to shape a product before launch rather than after the scandal

The most powerful intervention isn't a policy or a filter — it's a conversation without judgment.

Resources mentioned in this episode:

• Teens in the AI Era: Schoolwork and Skills That Matter — Common Sense Media
• Talk, Trust, and Trade-Offs: How and Why Teens Use AI Companions — Common Sense Media
• Youth AI Safety Institute and its free AI Risk Assessments of tools kids actually use
• Digital Literacy & Well-Being Curriculum, Grades 9–12 — free AI literacy lessons for teachers
• The Ghost in the Touchscreen: Social Scaffolds Promote Learning by Toddlers — Child Development, 2017
• Laura Zimmermann's research
• The Disengaged Teen — Jenny Anderson's episode, referenced on teen shame around AI use

Mentioned in this episode:

For those who wish to know more about the Education Futures fellowship: https://educationfutures.ai/fellowship