A parent-led exploration of youth + AI
Especially about what it could make possible for learning.
My tweens don’t have phones or social media. I’m cautious about introducing technology simply because it’s available, and I want a human in the loop while they learn to use AI.
At the same time, I’ve been exploring how AI can support my own productivity, learning and well-being.
My work in technology, privacy and AI governance made me eager to keep experimenting and learning, not only what AI can do, but how thoughtful ideas move into real-world adoption.
That professional lens made the questions at home harder to dismiss and more interesting to investigate.
My son had a math question. Through my account, I asked AI to guide him with questions rather than provide the answer. It also created the challenging practice problems he craved, extending beyond what I could confidently make myself.
AI expanded what I could provide as a parent. I was still the human filter.My daughter was overwhelmed by a martial arts belt exam. AI organized the material into a study structure and speech framework she still had to complete and practice herself.
AI supported the process. It did not manufacture the outcome.I went looking beyond my own experience. A nationally representative Common Sense Media survey of U.S. teens ages 13–17 gave me a much bigger classroom to look at.
Looking at the research through five questions.
AI is already part of how most teens approach schoolwork.
Base: teens age 13–17 (n=1,017).
Source: Common Sense Media, Teens in the AI Era: Schoolwork and Skills That Matter (2026). Nationally representative survey of 1,017 U.S. teens ages 13–17, fielded April 30–May 14, 2026. Some findings are reported among teens who use AI for schoolwork rather than among all surveyed teens. The dots are population-scale illustrations rounded to whole teens; they do not represent individual survey respondents, and a given dot should not be interpreted as the same teen from one lens to the next.
Kids are already using AI. Sometimes it supports the thinking; sometimes it shortcuts it. Rules and blocks can shape access, but they don’t teach a teenager the difference.
So where is AI actually reaching them?
03 · How AI reaches them
Some products are built for them directly. Others work through the adults around them. And some weren’t built for teens at all. Open a card for how it works and what it was rated.
Ratings come from the Common Sense Media Youth AI Safety Institute. A product-specific rating is used where available; otherwise the closest category assessment is labeled. Assessment dates matter because products change quickly.
Product ratings are one piece of a much bigger research picture. Across product safety, child development, family guidance, rights and governance, different organizations are studying different parts of the question.
Independent product testing, public risk assessments and evaluation methods built specifically around children and teens. The risk levels, dates and category ratings in section 03 are theirs, not mine.
13 assessments cited here Clinical partnerThe psychiatrists who co-authored the companion and mental-health assessments, testing across 13 clinical and developmental conditions.
United StatesA research ecosystem studying how AI affects learning over time, including methods for measuring outcomes beyond immediate task performance. OpenAI is also funding independent research into AI and teen development.
Overview ↗Updates ↗Teen research fund ↗Child-rights research and guidance on how AI and digital technologies affect young people, spanning responsible AI design, governance and emerging forms of technology-enabled harm. Its 2026 Through Children’s Eyes research also examines AI-generated imagery within the broader landscape of technology-facilitated child sexual exploitation and abuse.
GlobalAI for Children ↗Through Children’s Eyes ↗Human-rights review of chatbots used by 13- to 17-year-olds, with industry guidance in development.
Global · Meta-funded Clinical evidenceHealth advisories turning adolescent development evidence into AI-use recommendations.
United States Youth mental healthA 2026 study of nearly 40,000 students found that young people who used AI for emotional support were more likely to report emotional difficulties and loneliness. Researchers caution that this may be a signal of unmet need rather than evidence that AI caused the distress.
Canada · Research Family guidanceHow children actually use chatbots for learning, advice and companionship, including vulnerable children.
United Kingdom AcademicLong-running youth-and-technology research; a new AI and youth safety initiative began in 2025.
United States · HarvardMuch of the research and guidance referenced here comes from U.S. and U.K.-based organizations and should not be assumed to represent youth experiences globally.
Research is one part of the response.
Rules are another.Some efforts regulate AI systems directly. Others focus on online safety, companion chatbots or children’s rights. Together they create a growing set of protections, but no single playbook for responsible use.
Creates risk-based AI obligations and governance rules. Consequential for general-purpose AI, but not a child-specific framework.
The Online Safety Act already covers many chatbots used by children. New measures are extending protections around harmful features, emotional dependency, mental-health advice and age-appropriate use.
Requires AI disclosures, protections for known minors and protocols addressing suicide and self-harm interactions.
Includes federal youth online safety provisions and proposed chatbot safeguards, but has not completed the federal legislative process.
Ten requirements covering safety, privacy, fairness, transparency, children’s rights, development and AI literacy, supplemented by 2026 guidance specifically addressing AI chatbots and companions.
Chatbot policy brief ↗UNICEF, France, Spain and international partners are working to place children’s rights and well-being at the center of AI governance.
Rules can set boundaries.
Learning still happens in classrooms.Products can build safeguards. Governments can set rules. Schools are where those decisions meet the everyday reality of learning.
Access alone isn't the answer. How AI is designed, introduced and used can shape whether it supports the learning process or shortcuts it.
General-purpose AI can make students more successful at the task in front of them. But better performance while AI is available doesn't necessarily translate into learning once the tool is removed.
AI can be configured to preserve more of the student's thinking, guiding them through a problem rather than simply completing it.
AI doesn't have to sit between the student and the work. It can also support the educator, keeping human judgment and the learning objective at the center.
For kids with different learning needs, thoughtfully configured AI creates the possibility of support that adapts to the individual learner rather than assuming every student learns in the same way.
From purpose-built tutors to teacher-led classroom models and districtwide pauses, different approaches are testing where AI fits into the learning process.
Researchers tested AI-supported math instruction over two school years. Students made modest gains, but engagement with the AI tutor was lower than expected, suggesting that having access to a learning tool doesn't guarantee students will use it as intended.
A district initiative is focused on moving AI from policy into practice through staff training, classroom implementation and education for students and families.
Programs including ChatGPT for Teachers and new Khan Academy AI tools are expanding into more classrooms, giving educators and schools a more direct role in how AI is introduced and used for learning.
New York City and Los Angeles are taking more restrictive approaches to student-facing generative AI for the 2026–27 school year. NYC is pausing it through eighth grade while allowing limited, vetted high-school programs. LAUSD has blocked student access on district-issued devices while it develops new guidance and AI-literacy curriculum.
Different approaches are already being tested.
A pattern is starting to emerge.Across the research and experimentation, four conditions keep surfacing.
Design choices shape whether AI behaves like an answer engine or a learning scaffold.
Educators bring context about the student, the assignment and what the work is meant to develop.
Support can enter when a student is genuinely stuck without removing every chance to struggle and try.
Responding to individual learning needs rather than assuming every student needs the same support.
The opportunity may not be more AI. It may be better orchestration of when, how and why AI enters the learning process.
I'm the one deciding when AI might be useful, how it enters the conversation and making sure it helps my kids work through something rather than simply doing the work for them.
It's a very human system. And a very temporary one.As my kids get older, I won't always be the one deciding how and when AI enters the picture. The opportunity isn't to recreate a parent as the guardrail.
It's to figure out which parts of that judgment technology can support, which belong with educators and families, and which young people ultimately need to learn to exercise for themselves.Recent developments I'm watching across youth, AI and learning.
Curated, not comprehensive · Reviewed 8 Sept 2026
I'm a San Diego-based brand and marketing communications leader with 15+ years of experience helping organizations translate complex ideas into clear, compelling narratives for Fortune 500 brands and high-growth technology companies.
Most recently, I worked at Transcend, a data privacy and AI governance technology company, where my work crossed brand, communications, product marketing and emerging AI initiatives. I'm also a mom of two tweens, which is where the professional and personal sides of this rabbit hole collided.
I started exploring this topic while between roles, partly because I couldn't stop thinking about it and partly because I wanted to get hands-on with some of the newer AI products changing how websites are built. One question led to another, and the rabbit hole got considerably deeper.
I'm not a child development or education expert. This is a personal, evidence-informed exploration shaped by my professional experience and my perspective as a parent.
I'm interested in how my background might support the people already doing this work: helping translate research into something useful, making complex questions around youth and AI easier to understand, and bringing thoughtful initiatives from idea to adoption.
“Denise has a real talent for bringing clarity and structure to complex, ambiguous work. She’s thoughtful, collaborative, and exceptionally reliable, with an ability to take an idea and move it toward execution.”
Kate ParkerPresident, Transcend
“She was exceptional at leading complex, cross-functional initiatives: setting strategy, aligning stakeholders, and moving projects forward without losing sight of the details. She has an incredibly high bar for quality and a sharp eye for the small things that can make or break execution.”
Morgan SullivanMarketing, Transcend
“Denise is awesome! She brought so much curiosity, thoughtfulness, and enthusiasm to our work at Transcend. I especially appreciated her eagerness to learn about AI and explore practical ways it could support our work. She helped create tools that made technical concepts easier for customers to understand. … She has a wonderful ability to balance the details of a project with the bigger picture of the brand and customer experience.”
Kat Mena HarrisAI Strategy & Strategic Ops, Transcend
“She was a phenomenal teammate on all fronts — pushing forward strategy, competitive research, design partnerships, and more. Her energy and instinct for market trends was an enormous asset to any team working to build in a cutting-edge space.”
Rowan StewartAI & Data Safety Product Leader
“Denise is everything you want as a client in an agency partner — religiously across every last detail, accommodating of developments on our side yet firm with deadlines and requirements, and a joy to work with. Crucially, she gave me confidence from the outset that the project was on track at all times, and helped us exceed our expectations.”
Andrew MoonBrand & Comms, TranscendFormer client · later manager at Transcend