Research institute · Common Sense Media
Two-thirds of US teens use AI, parents 18 points behind
Common Sense Media's late-2025 nationally representative survey of US parents and 12 to 17 year-olds maps how children and AI now meet at home. Two-thirds of teens use AI tools regularly, parents trail at 49%, and both groups want firmer rules even as they split on whether the technology helps or harms.
Source published Summarised
Key findings at a glance
- 67% of US kids and teens age 12 to 17 use AI chatbots, companions, programs or tools often or sometimes. Two-thirds is the new floor for teen AI usage.
- 49% of US parents use AI often or sometimes. The 18-point usage gap with their kids is the central misalignment in the data.
- 65% of parents think AI will change daily life as dramatically as the internet or electricity. Families see the technology as foundational, not novel.
- 71% of parents agree people will be so dependent on AI they will not be able to function without it by the time today's kids are adults.
- 60% of kids and teens agree on the same dependence point. Both cohorts share the worry even where their optimism otherwise diverges.
- 57% of parents of under-18s think AI will make it harder for their child to find a job. Only 29% think it will make it easier.
- 58% of kids and teens say AI will have a positive impact on their school learning. Only 41% of parents share that view.
- 84% of parents are concerned about AI misusing their child's personal information. The single biggest worry in the parent survey.
- 58% of parents know little or nothing about the safety features in AI products their teen might use. The information gap mirrors the usage gap.
- 75% of parents support establishing a government oversight body for AI safety. 72% of kids and teens agree, and support is bipartisan in the US sample.
What the report covers
Common Sense Media commissioned two parallel nationally representative US surveys in late November 2025: one of 1,244 parents (weighted base 843) and one of 1,100 kids and teens age 12 to 17 (weighted base 800). Fieldwork ran 17 to 26 November for parents and 17 to 28 November for kids and teens. Both were online surveys conducted by Lake Research Partners with Echelon Insights.
The questionnaire covered five threads: how families use AI today, how they expect it to affect society and jobs, how AI is turning up in school work, what parents know about safety controls, and what regulation, if any, families want. Some questions were split-sampled to control for wording effects across demographic, regional and partisan groups.
What the data captures is generational tension over one technology. On the subject of children and AI, kids and teens are more likely to use the tools, more likely to treat them as a learning asset, and more optimistic about the long-term payoff. Parents are more cautious on jobs, on ethics in school work, and on the gap between what AI promises and what it delivers. Both groups converge sharply on safety, privacy and the appetite for hard rules. The pattern echoes the parent-versus-teen split in Pew's survey of how US parents view teen social-media use, where parents skewed 3-to-1 negative on platform impact while teens sat closer to neutral, and it tracks the platform-by-platform behaviour mapped in our summary of how teens use TikTok, Instagram and Snapchat differently.
What this means for UK marketers
The data is US-only, but the UK behavioural parallel on every comparable consumer-tech metric Ofcom and the IAB UK have published in the last 18 months is close enough to read this as a leading indicator for British family behaviour. Four operational lessons follow for UK marketers.
Brands selling to families need to show, not tell, on data and safety. 84% of US parents are concerned about AI misusing their child's personal information. Generic "AI-powered" or "smart" product copy now reads as a risk flag in this segment, not a differentiator. UK marketers in EdTech, parenting, family services, kids' apps and family-facing ecommerce should expect to spend more brief time on the trust mechanics (what data is collected, where it goes, age-gating, opt-outs) than on the AI capability itself. The ICO Children's Code already sets the legal floor in the UK, and consumer expectation now runs ahead of it.
EdTech and learning brands have a parent-versus-kid pitch problem. The 17-point gap between teens (58%) and parents (41%) on whether AI helps learning means the buyer and the user read the same product through different lenses. Family-targeted creative needs two copy paths: a parent-facing message about responsible adoption, oversight, age-appropriate use and learning outcomes, plus a kid-facing message about productivity, autonomy and creativity. This is not a tone problem, it is a benefit-architecture problem, and it shapes how a brand should split paid acquisition budget across audiences and creatives.
The unanimity on guardrails is a regulatory weather signal. 75% of parents and 72% of kids back a government oversight body for AI safety, with bipartisan support in the US sample. The UK already has the Online Safety Act and the ICO's Children's Code in force, and Ofcom's online-safety regime is tightening the rules on services children use. Brands building AI into family-facing products should expect explicit disclosure norms, mandatory age verification and verifiable claims to become the default within the next 24 months.
One under-priced channel implication: parents sit 18 points behind their kids on AI usage yet are heavily concerned about how those kids use it. That is a clean targeting brief for Meta-led parent-education campaigns, where a brand earns trust by helping parents close the knowledge gap rather than selling around it. The wider context is a fast-growing domestic AI economy: the UK now hosts thousands of AI firms turning over billions, as our note on the DSIT UK AI sector data sets out. Brands that pre-empt the regulatory direction with clear product labels, transparent data flows and credible third-party assurances will not only avoid future fines, they will beat the segment median on conversion in any category where parents are the gatekeeper.
How they got the data
Two parallel online surveys conducted in late November 2025. The parent survey, fielded 17 to 26 November 2025, reached 1,244 US parents nationwide, with a base sample of 843 and oversamples of 100 Black, 100 Latino, 101 Asian American and Pacific Islander parents, and 100 parents of children age 12 to 17. The kid and teenager survey, fielded 17 to 28 November 2025, reached 1,100 US kids and teens age 12 to 17, with a base of 800 plus oversamples of 100 Black, 100 Latino and 100 Asian American and Pacific Islander respondents. All oversamples were weighted back to their proper proportion of the universe.
Lake Research Partners and Echelon Insights designed and administered both surveys for Common Sense Media. Some questions were split-sampled to test wording effects across demographic, regional and partisan groups. Quantitative results are expressed as percentages and may not sum to 100% due to rounding or multiple responses. The data is US-only.
Research institute
Read the full report
From Common Sense Media · published 09/03/2026