Quick safety summary
Machine Learning for Kids at a glance
85/100
kidsafe
Ages Ages 6–16 (teacher-managed)
Within intended range
High
Score 90/100
82/100
Strong filters
Best use cases
- AI / ML Education Platform (K-12) workflows
- Classroom assignments
Main risks
- No elevated risks identified.
Recommended for
Not ideal for
Product overview
How Machine Learning for Kids works
Machine Learning for Kids is a free, teacher-led platform that lets children build simple machine-learning models (image, text, sound, and number classifiers) and use them in Scratch or Python projects. Created by Dale Lane and supported by IBM, it is one of the safest AI tools for children currently available: a teacher-managed account model eliminates the need for children to hand over personal information, there is no chatbot, no image generation, no advertising, and no commercial upsell. It is widely used in UK primary and secondary schools, CoderDojo, and Code Club, and is a strong fit for AI literacy under the EU AI Act's emphasis on transparent, educational AI experiences.
Free full report — 0 of 3 used
You can open 3 complete evaluations on the free plan. Already-opened reports stay unlocked. Upgrade to School Pro for unlimited full reports and the K-12 governance workspace.
Risk intelligence
Risk level breakdown
Risk dimensions derived directly from the SafeGradeAI rubric scores (higher score = higher risk). Additional dimensions appear when entered by an evaluator.
Scores derived from the SafeGradeAI rubric. Methodology: read more.
Safety framework
Framework analysis
Five evaluation categories from the SafeGradeAI rubric. Each shows the rubric score, evaluator confidence, and (where documented) analysis, strengths, and concerns.
Strengths & concerns
Detailed pros & concerns
Strengths
- Teacher-managed account model — children do not submit personal information at signup
- No chatbot, no generative image or text, no user-to-user messaging, no public feed
- Free to use, ad-free, no upsell, no premium tier marketed to children
- Curriculum-aligned worksheets and lesson plans provided by the author
- Widely used and reviewed in UK schools, CoderDojo, and Code Club
- GDPR / UK-GDPR-facing operator based in the United Kingdom
- Supports AI literacy explicitly named as a priority under the EU AI Act
- Underlying ML services (IBM Watson) run under IBM enterprise terms, not consumer data-sharing
Concerns
- Not designed for unsupervised home use — assumes a teacher or parent leader
- No independent third-party child-safety certification listed (e.g., kidSAFE, iKeepSafe)
- Content safety of training data depends on teacher oversight of student uploads
- Depends on continued availability of IBM Watson tiers used behind the scenes
- US-facing FERPA documentation is thinner than large commercial K-12 vendors
- Interface is functional rather than polished; younger children need adult support to navigate
Guidance at a glance
Recommendations for parents, teachers, and administrators
Parent guidance
Machine Learning for Kids is one of the AI tools we most comfortably recommend for children — but it is designed for supervised, project-based learning, not casual home use. There is no chatbot to talk to your child, no images generated for them, and no attention economy pulling them back. If your child is homeschooled or you want to teach them ML concepts, create a teacher/leader account yourself and sit alongside them; the platform's tutorials are the intended entry point. Children under 13 can safely use it in this teacher-led configuration because the platform never asks children to submit their own personal information.
Classroom-use recommendation
Strongly recommended for classroom AI-literacy units grades 3-10; students never create accounts and all projects live under the teacher's login. Pair each activity with an explicit discussion of training data, bias, and the "black box" so it builds critical AI understanding rather than novelty.
School policy recommendation
This is a strong fit for K-12 classroom deployment. The teacher-managed account model, closed content surface, GDPR/UK-GDPR posture, and clear educational purpose align well with FERPA, SOPPA, SOPIPA, NY Ed Law 2-d, California AB 1584, and comparable state student-data-privacy frameworks — schools should still request or reference the operator's privacy policy in their district AI-tool review, and record the underlying IBM Watson processing in their data map. Under the EU AI Act, this is a limited-risk educational AI system supporting the Act's AI-literacy objectives; EU schools should still complete their own DPIA. Multiple US states now require formal district AI policies (e.g., Ohio's July 1, 2026 mandate); Machine Learning for Kids is a defensible entry on an approved-tool list.
Trust & credibility
Evaluation integrity
Last reviewed
July 15, 2026
Report version
v1
Evaluator confidence
86/100
Monitoring
Active continuous monitoring
Methodology
Scored against the SafeGradeAI rubric covering privacy, moderation, generative safeguards, age-appropriate design, and transparency.
Read full methodology →Change log
- July 15, 2026Evaluation published.
Re-tested at least every 6 months and after any major release.
What changed since the previous evaluation
Reviewed July 15, 2026 · Confidence: high
- 2017: Machine Learning for Kids launched by Dale Lane
- 2018-2020: Adopted broadly by CoderDojo, Code Club, and UK primary and secondary schools
- 2022: Python integration expanded alongside existing Scratch integration
- 2024: EU AI Act published; the platform aligns with its AI-literacy objectives
- 2025: Worksheets and lesson-plan library refreshed
Sources & citations
Every major claim in this evaluation is backed by at least one credible public source. Last reviewed July 15, 2026.
- 1Machine Learning for Kids — homeDale Lane · 2025-09-01 · machinelearningforkids.co.uk
- 2About Machine Learning for KidsDale Lane · 2025-09-01 · machinelearningforkids.co.uk
- 3Worksheets and lesson plansDale Lane · 2025-09-01 · machinelearningforkids.co.uk
- 4Raspberry Pi Foundation — CoderDojo & Code Club (uses ML for Kids)Raspberry Pi Foundation · 2024-06-01 · raspberrypifoundation.org
- 5IBM Watson Assistant — underlying ML serviceIBM · 2025-01-01 · ibm.com
- 6ICO — Children's data and UK GDPRUK Information Commissioner's Office · 2024-01-01 · ico.org.uk
- 7Children's Online Privacy Protection Rule (COPPA)US Federal Trade Commission · 2013-07-01 · ftc.gov
- 8What is FERPA?US Department of Education · 2024-01-01 · studentprivacy.ed.gov
- 9EU AI Act — Regulatory frameworkEuropean Commission · 2024-08-01 · digital-strategy.ec.europa.eu
Methodology
How this score was produced
Scored against the SafeGradeAI framework v2.5 across five dimensions — privacy, moderation, GenAI safeguards, age-appropriate design, and transparency — with COPPA, FERPA, and GDPR-K lenses. Independent. No vendor pay-for-rating.
Independent evaluation. SafeGradeAI does not accept payment from vendors for ratings.
Official website
Visit Machine Learning for Kids directly
For the most current product capabilities, pricing, and policies, visit the official Dale Lane (independent, IBM-supported) website.
machinelearningforkids.co.uk