United States · Grades 9 to 12 · AI and machine learning
AI courses for high school students
An American teenager who wants to learn AI properly has more choice than ever, and several of the strongest options are free. There is a full-year school course written for high schools, a university course from Harvard that anyone can follow, a crash course from Google, and live classes with a teacher. They differ in one way that matters more than price: whether the student ends up understanding how a model learns, or only how to use one. This page sets out what each kind of course actually covers, using the providers' own descriptions, so a family can choose with open eyes.
Live teaching since 2020 · 10,000+ students · AI taught with code, not slides
The short answer
High school students in the US can take AI at school through CodeAI's AI Foundations (up to 180 class periods, builds toward AP CSP or AP CSA), free online through Harvard's CS50 AI (seven weeks, needs about a year of Python) or Google's Machine Learning Crash Course, or live with a teacher. Modern Age Coders teaches AI and Python live online for teenagers 13 to 18, in a group batch or one to one.
Learn it live
Teen courses that run from a first line of code to a trained model
These are the three we usually suggest first to a high schooler. A regular teacher runs each one live.
These are categories, not a ranking. Each row names one real example and describes it from the provider's own page, read on 17 October 2026.
AI courses open to US high schoolers, by kind (examples described from the providers' pages, read 17 October 2026)
Kind
Example
Length
Needs first
Live teacher?
School course
CodeAI Artificial Intelligence Foundations
90 class periods per semester, 180 for the year
No prior experience
Your school's teacher
Free university course
CS50's Introduction to AI with Python
Seven weeks of material
CS50x or a year of Python
No; recorded lectures
Free industry course
Google Machine Learning Crash Course
12 self-contained modules
Algebra helps (our view)
No; self-paced
Live classes
Teen AI and Python courses, like ours
Months, one class at a time
Nothing, for a Python start
Yes, every class
How the four fit together
They are not rivals. A student might take AI Foundations at school in grade 9, learn Python properly in grade 10, and work through CS50 AI the summer after. Each kind covers ground the others assume.
The gap to plan for is the year of Python between a school course and a university one. Harvard's course page asks for CS50x or at least one year of experience with Python before starting.
Where a live teacher earns its place
Free courses are excellent and we recommend them. What they cannot do is notice that a student copied a solution without understanding it, or explain the same idea a third way.
A live class is worth paying for when a student has stalled, needs a pace set for them, or wants feedback on their own project. If none of those is true, start free.
AI Foundations: a free course US high schools can offer
The nonprofit long known as Code.org now presents itself as CodeAI; its own footer says "Code.org" is a former trademark of CodeAI. Its high school AI course is free to schools, and a student can ask a teacher or counselor whether it is on the schedule.
CodeAI Artificial Intelligence Foundations, from the curriculum page (read 17 October 2026)
Point
What CodeAI says
Length
90 class periods per semester (180 total) for the full year; semester and unit options too
Prior experience
No prior experience required
Leads to
Builds toward AP CSP, AP CSA, or workforce readiness
Standards
Aligned to CSTA 2026 standards and the TeachAI AI Literacy Framework
Tools
AI Chat Lab, Python Lab, Web Lab, Sketch Lab
Devices
Chromebook or computer
A Python unit
Foundations of AI Programming: "Learn Python through hands-on projects"
State versions
Versions specifically designed to meet Idaho standards and Indiana standards
If your school offers it
Take it. A full year is 180 periods, which is 180 / 2 = 90 a semester, a lot of structured AI time, and CodeAI lists Chromebooks among its supported devices.
Use the second semester well. CodeAI says students learn JavaScript and build web apps with AI as a creative partner. Ask the teacher how much code students write themselves, and write more at home.
If it does not
Ask. Curriculum choices are made by schools and districts, so a request from interested students is a reasonable place to start.
In the meantime the same ground can be covered outside school: Python first, then a structured AI course. That is the route the rest of this page describes.
Free online
Two free AI courses worth a high schooler's time
Both are free to read and follow, and both are written by people who teach this material for a living. Neither has a teacher who will check in on your teenager, so they work for self-driven students, ideally paired with some feedback.
CS50 AI week by week, from the course page (read 17 October 2026)
Week
Topic
In one line (ours)
0
Search
How a program finds a path, such as the way out of a maze
1
Knowledge
Logic a computer can reason with
2
Uncertainty
Probability, when the answer is not sure
3
Optimization
Picking the strongest option among many
4
Learning
Machine learning from examples
5
Neural Networks
Layers of simple units that learn together
6
Language
How programs process human language
CS50 AI, honestly
Harvard describes the course as seven weeks of material, each with a lecture and a project, and anyone can take it free through its OpenCourseWare. Its weeks run Search, Knowledge, Uncertainty, Optimization, Learning, Neural Networks and Language, and the projects are Python programs the students write themselves, which is what makes it valuable.
It is not a first course. The prerequisite is CS50x or at least one year of experience with Python, and submitting projects for feedback needs an edX account, and the edX terms set a minimum age of 13.
Google's Machine Learning Crash Course
Google says that since 2018, millions of people worldwide have relied on Machine Learning Crash Course. It is organized as 12 self-contained modules in four groups: ML models, Data, Advanced ML models and Real-world ML.
The modules include Datasets, generalization, and overfitting, Intro to Large Language Models, and Fairness. Google describes the format as animated videos, interactive visualizations and hands-on practice exercises.
Federal policy now points the same way. An executive order signed on April 23, 2025 asks the agencies on its AI education task force to develop online resources focused on teaching K-12 students foundational AI literacy and critical thinking skills, and tells the Secretary of Labor to support the creation of opportunities for high school students to take AI courses and certification programs. Free, serious courses like these two are what that looks like in practice.
A first model, in real code
What lesson one of a real AI course looks like
Every serious AI course starts with the same idea: train on some data, test on data the model has not seen. Here it is in 20 lines, using the iris flower dataset that ships with scikit-learn. The script is ours, tested on 17 October 2026 under scikit-learn 1.7.2 and Python 3.13.
Choosing k for a nearest-neighbor modelpython
# Lesson 1 of an AI course, in code (ours)
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import accuracy_score
iris = load_iris()
X_train, X_test, y_train, y_test = train_test_split(
iris.data, iris.target, test_size=0.3,
random_state=42, stratify=iris.target)
for k in [1, 3, 5, 15, 45]:
model = KNeighborsClassifier(n_neighbors=k)
model.fit(X_train, y_train)
guess = model.predict(X_test)
print(k, round(accuracy_score(y_test, guess), 3))
flower = [[5.9, 3.0, 5.1, 1.8]] # cm: sepal, sepal, petal, petal
chosen = KNeighborsClassifier(n_neighbors=5).fit(X_train, y_train)
print(iris.target_names[chosen.predict(flower)[0]])
Output on our run: k=1 0.933, k=3 0.956, k=5 0.978, k=15 0.956, k=45 0.889, then virginica for the example flower. The dataset has 150 flowers, so test_size=0.3 holds back 150 x 0.3 = 45 of them for testing.
What the student should notice
Accuracy rises from k=1 to k=5 and then falls. With k=45 the model asks 45 neighbors to vote, nearly half the training flowers, and the answer blurs. That rise and fall is the first lesson in model selection.
The test flowers were split off before training and stratified, so each species appears in the same proportion. A course that skips this step is teaching a habit that breaks real projects.
What a good course does next
It warns that choosing k by looking at the test score is a quiet form of cheating, and introduces a separate validation set or cross-validation to fix it.
Then it moves from flowers to the student's own data. scikit-learn's documentation lists the iris set as 150 samples with 4 features; a student's first real dataset will be messier, which is the point.
Choosing
Five questions that tell you what an AI course really teaches
Our checklist for reading any AI course page, free or paid, ours included. Course names are cheap; these questions get to what a student will actually be able to do.
Does the student write the code? Look for projects in Python, not only demonstrations of apps. A course where AI writes everything teaches prompting, which is a different skill.
Is there a test set? Any syllabus that trains models should mention testing on unseen data, and ideally overfitting by name.
Who checks the work? Recorded courses rely on automated checks or none. Decide whether your teenager needs a human to read their code.
Are the tools allowed for their age? Several popular AI chatbots set adult age limits in their terms. A course for under-18s should use tools whose terms allow teenagers.
What exists at the end? A finished project the student can explain is the only outcome worth paying for. Ask to see one from a past student.
Running the five questions on the four kinds of course (our judgment)
Our suggested sequence for a student who starts with no code in grade 9. It mixes school, free and live courses; swap in whatever your school offers.
AI through grades 9 to 12 (our plan)
Grade
Course
By the end of the year
9
AI Foundations at school, or Python live
Writes small Python programs; knows what training data is
10
Python properly, plus a first machine learning course
Has trained and tested a model like the one above
11
CS50 AI or Google's crash course, with feedback
Can explain search, probability and neural networks simply
12
A project of their own, with a teacher reviewing
Has one finished project with an honest write-up
Starting late
A junior with no code can compress grades 9 and 10 into one year of Python and machine learning. It is a busy year, not an impossible one.
Starting early
A middle schooler who already codes can begin at grade 10 of this path. Depth matters more than finishing early.
AP alongside
AI Foundations says it builds toward AP CSP or AP CSA. Our pages on AP CSP and AP CSA cover those exams.
From user to builder
Four stages of a high schooler learning AI
Where a teenager might start, and which kind of course moves them on.
Stages of learning AI in high school (ours)
Stage
Sign of it
1. User
Uses AI apps every day; cannot say how they work
2. Coder
Writes Python; has not yet trained a model
3. Trainer
Trains and tests models; reads accuracy honestly
4. Builder
Chooses a problem, a model and a test, and explains all three
For parents
The jump from stage 1 to 2 is the one to check a course does not skip. A course that goes straight to AI apps leaves a teenager at stage 1 with better vocabulary.
Ask your teenager to show you a test score and explain it. That conversation tells you their stage.
For students
Learn Python before you learn AI. Every course at stage 3 assumes it.
Change one setting at a time and write down what happened. That habit is most of machine learning.
The course catalog
Nine courses for high schoolers learning AI
Live and online, with the same teacher each week, in a batch or one to one.
A weather forecasting site with live conditions for any location.
by Krish
Fees
Fees
School and free courses cost nothing, and we say so above. Our live classes are billed monthly in US dollars, with no semester commitment, and you can stop after any month.
"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."
Ria Mukherjee
Parent
★★★★★
"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."
Samriddha Mondal
Student
★★★★★
"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."
Vansh Agarwal
Student
★★★★★
"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."
Sonam Oswal
Parent of Dhairya
★★★★★
"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."
Sonu Goyal
Parent
★★★★★
"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."
Ritu Kedia
Parent
Questions from families
AI courses for high schoolers: common questions
What AI courses can high school students take?
Four kinds: a school course such as CodeAI's Artificial Intelligence Foundations, a free university course such as Harvard's CS50 AI, a free industry course such as Google's Machine Learning Crash Course, and live classes with a teacher.
Are there free AI courses for high school students?
Yes. CS50 AI can be followed free through Harvard's OpenCourseWare, Google's Machine Learning Crash Course is free to read online, and CodeAI offers AI Foundations to schools at no cost.
Is CS50 AI good for a beginner?
Not as a first course. Harvard lists the prerequisite as CS50x or at least one year of experience with Python. Learn Python first, then take it.
How long is the AI Foundations course?
CodeAI lists 90 class periods per semester, 180 for the full year, with shorter semester and unit options. It needs no prior experience.
Does AI Foundations count toward AP?
CodeAI says it builds toward AP CSP, AP CSA, or workforce readiness. It is not itself an AP course; AP exams are run by the College Board.
What age do you need to be for CS50 AI?
Following the free materials needs no account. Submitting projects for feedback needs an edX account, and edX's terms say you must be at least 13 years old to use its service.
Should a high schooler learn Python before AI?
Yes, for anything beyond using AI apps. The university and industry courses on this page assume Python, and real machine learning projects are written in it.
Do your AI courses use ChatGPT?
For under-18 students we choose tools whose terms allow teenagers, and we teach students to check those terms themselves. The core of every course is Python code the student writes.
When are classes for US high schoolers?
Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays.
How much do your AI classes cost?
The first class is free. After that, families pay monthly for live teaching, and the fees section of this page shows the figure in US dollars.
We teach from India on live video, which is why there is no local center to visit and why any US time zone works. WhatsApp reaches us fastest.
Want a class sooner? Book a priority demo
A priority demo is a full class of about 45 to 60 minutes. After the class the mentor sends a written skill report and a personal learning roadmap on WhatsApp. If you enrol, the demo fee is adjusted against your first month's fee. Share your demo Order ID with us on WhatsApp and we take it off.
