AIMS / IDE / 01
AI literacy
Understanding AI, not just using it.
Open the syllabusUnited States · Grades 6 to 8 · Ages 11 to 14
Middle school is when AI stops being a toy and starts being homework help, a search engine and a source of rumors, all at once. It is also, conveniently, the age when students can understand how it works: that a model learns from examples, that the examples can be skewed, and that a confident answer can be wrong. US schools have started teaching exactly this. This page explains what that school teaching now covers in grades 6 to 8, what a good outside class should add, which tools suit an 11 year old versus a 14 year old, and what a first model looks like in a few lines of Python.
Live teaching since 2020 · 10,000+ students · AI explained to the age, not dumbed down
The short answer
A good AI class for middle schoolers teaches how models learn from data, where they go wrong (bias, made-up answers), how to check AI output, and a first model in code. US schools increasingly cover this through CodeAI's AI Discoveries and Day of AI. Modern Age Coders teaches AI live online for grades 6 to 8 through its AI Literacy course (ages 8 to 14) and its Python courses, as a batch or one to one.
Live AI classes for grades 6 to 8
Chosen for ages 11 to 14. Each is taught live by a regular teacher who adjusts the pace to the student.

AIMS / 01
How AI learns, where it fails, and how to check it, with small builds.
Open the syllabus →
AIMS / 02
Real typed code for sixth and seventh graders, with smart projects.
Open the syllabus →
AIMS / 03
The teen Python course, for students ready for longer programs.
Open the syllabus →What school now offers
Two nonprofits now publish full AI courses for middle school, free for any school to adopt. Knowing what they cover helps a parent judge what an outside class should add. Both are described here from their own pages, read on 17 October 2026.
| CodeAI AI Discoveries | Day of AI PreK to 12 curriculum | |
|---|---|---|
| Who publishes it | CodeAI, the nonprofit formerly presented as Code.org | Day of AI, founded in 2021 through MIT's RAISE initiative |
| Middle school content | A unit on AI and machine learning: how ML works, training data, classification models, bias in AI systems | Grades 5 to 8: predictions, data, machine learning, bias, fairness and privacy |
| Length | A quarter, a semester or a school year | Grade-by-grade lessons across the year |
| Also covers | Web development, games, apps, data, devices | Fact-checking AI output, data privacy, environmental impacts |
| Released | Current course page | Announced August 19, 2026 |
| Cost to schools | Free | Free to educators, schools and districts |
CodeAI rebuilt its middle school course around AI. Its page says AI and Machine Learning moves to Unit 2, so AI ideas come early in the year rather than at the end, and AI Discoveries connects directly to AI Foundations, the high school course.
The physical computing unit lets classes choose between Circuit Playground or BBC micro:bit, and an AI Tutor in programming units helps students work through problems without giving away the answer.
Day of AI says it became an independent nonprofit in 2024 and has reached over 4 million students. Its new curriculum, announced in a release dated August 19, 2026, is built as a progression from preschool to grade 12.
For grades 5 to 8 the release lists predictions, data, machine learning, bias, fairness, and privacy, and among its broader topics, mandatory fact-checking of AI-generated content. That last one is the habit most worth bringing home.
Sources, read 17 October 2026: CodeAI AI Discoveries; the Day of AI curriculum release. We are not connected with CodeAI or Day of AI. Ask your child's teacher which, if either, the school uses.
What to learn
Our list, built from what the school curricula above emphasize and from the questions middle schoolers bring to class. A class that covers only the first row is a tool tutorial, not an AI class.
| Topic | What the student can do after | A question that tests it |
|---|---|---|
| Using AI carefully | Ask a clear question and judge the answer | "How would you check that?" |
| Learning from examples | Explain that a model copies patterns in its training data | "Where did it learn that?" |
| Bias and gaps | Spot when examples leave someone or something out | "Who is missing from the examples?" |
| Wrong but confident | Know that fluent text can be invented | "Is there a source for this?" |
| A first model in code | Train and test a tiny model themselves | "What happens if we add one more example?" |
Sixth graders meet AI in search results, recommendations and voice assistants, often without knowing it, and chatbots are a click away by eighth grade. Learning how these systems work before habits set is easier than unlearning them at 16.
Middle schoolers are also ready for the core idea. A twelve year old can follow why a model trained only on loud songs will call every song loud, and that one example carries most of AI literacy.
A coding class teaches a language. An AI class teaches how systems learn from data, using code as one of its tools. The strongest mix is a class where students write the small model themselves.
For general programming at this age, from Scratch to Python, see our page on coding for 12 year olds. This page stays on AI.
Safe tools by age
Grades 6 to 8 straddle an important line: 13 is a common minimum age in online terms (edX, for one, requires users to be at least 13), and some services set a higher age. Read each tool's own terms before a child signs up. These are the points we could check at source.
| Grade | Typical age | What to know |
|---|---|---|
| 6 | 11 to 12 | CodeAI says Web Lab and AI tools are built for students ages 11 and up |
| 7 | 12 to 13 | Machine Learning for Kids asks under-13s to use only a small language model with an adult supervising |
| 8 | 13 to 14 | Some services open at 13, some only with a parent's consent, and some later; check each one's terms |
| Any | Any | Plain Python on the family computer needs no account at all |
CodeAI states that AI Tutor interactions in AI Discoveries are stored securely and automatically deleted after 90 days, and that student interactions are never used to train third-party AI models. That is the kind of statement to look for in any tool a school uses.
Machine Learning for Kids, a free teaching tool, tells users: "If you are under the age of 13, please only use a small language model with supervision from a trusted adult." It is a sensible rule well beyond that one site.
For students under 13 we teach the ideas with Scratch, with tools built for children, and with Python that runs offline. Nothing needs the child to create an account on an adult service.
For 13 and 14 year olds we add tools whose terms allow teenagers, with a parent's consent where the terms ask for it. Our AI tool age guide lists what popular tools' terms say.
For homework, the line that matters is honesty about what the student did. Our guide to safe AI tools for homework goes through it, and teaching kids AI at home has activities that need no tool at all.
A first model
This is the first model we suggest a middle schooler writes. It uses plain Python, no libraries to install and no accounts, and it learns from eight examples. We wrote it and checked the output under Python 3.13 on 17 October 2026.
# Teach the computer: dance song or chill song? (ours)
import math
# examples: (beats per minute, loudness 1 to 10) -> label
examples = [
((128, 8), "dance"), ((124, 9), "dance"),
((118, 7), "dance"), ((132, 8), "dance"),
((70, 3), "chill"), ((82, 4), "chill"),
((90, 2), "chill"), ((76, 5), "chill"),
]
def guess(song):
# find the example closest to the new song
nearest = min(examples,
key=lambda ex: math.dist(ex[0], song))
return nearest[1]
print(guess((120, 6))) # a new song
print(guess((95, 3)))
print(guess((100, 6))) # a hard oneOutput on our run: dance, then chill, then chill. Python's math.dist gives the Euclidean distance between two points, so the model simply finds the example song closest to the new one and copies its label.
A song at 100 beats per minute and loudness 6 sounds like it could go either way. The model says chill. Our arithmetic: its distance to the chill example (90, 2) is about 10.8, and to the dance example (118, 7) about 18.0.
Why? Tempo numbers run from 70 to 132, while loudness only runs from 2 to 9. A difference of 18 in tempo outweighs everything loudness can say. The model is not weighing the two features fairly.
That a model only knows its examples, and that the way data is measured changes the answer. Real machine learning engineers fix this with scaling, and a middle schooler has now seen why it matters.
Next steps in class: add examples and watch answers change, invent a song the model gets wrong, then rescale tempo to 1 to 10 and run it again.
Twelve weeks
Our plan for a first term, for students in grades 6 to 8 with or without coding. Students who already code move faster through weeks 1 to 4.
| Weeks | Focus | Made by the student |
|---|---|---|
| 1 to 2 | What AI is and is not; where they already meet it | A list of AI they used this week, sorted by how sure they are |
| 3 to 4 | Examples and patterns, in Scratch or on paper | A sorting game a friend can play |
| 5 to 6 | A first model in Python | The dance or chill script, extended with their own songs |
| 7 to 8 | Bias and missing data | A model that fails on purpose, and a fix |
| 9 to 10 | Checking AI answers | Three AI claims, each checked against a real source |
| 11 to 12 | A small project of their choice | A model, a test and a two-minute explanation |
More discussion and Scratch, shorter Python. The ideas are the same; the code is lighter.
The full plan as written, with the project in weeks 11 and 12 chosen by the student.
Faster through the basics, then a second model with a real dataset and an honest test.
Schools are adding AI activities too. CodeAI says the Hour of Code is now the Hour of AI, with 100 one-hour tutorials available in over 45 languages. A one-hour activity is a good taster; a term of classes is where understanding sets in.
From user to maker
Where a sixth grader might start, and where a year of good teaching can take them.
| Stage | How you can tell |
|---|---|
| 1. User | Uses AI apps; trusts most answers |
| 2. Questioner | Asks where an answer came from, and checks |
| 3. Explainer | Can say how a model learns from examples |
| 4. Maker | Has written a small model and broken it on purpose |
Stage 2 is the one that protects your child. Ask "how would you check that?" at dinner and you are teaching it.
Stage 4 is where a middle schooler becomes genuinely ahead: not using more AI, but understanding it.
When an AI answer surprises you, ask where it learned it. That question is most of AI.
Write your own tiny model. Once you have, every AI app makes a bit more sense.
The course catalog
Live online with a regular teacher, as a batch or one to one. Courses marked 13 and up suit eighth graders who have turned 13.
AIMS / IDE / 01
Understanding AI, not just using it.
Open the syllabusAIMS / IDE / 02
Patterns and logic, before typed code.
Open the syllabusAIMS / IDE / 03
Ages 10 to 15: the math under the models.
Open the syllabusAIMS / COD / 01
Games, turtle art and first smart programs.
Open the syllabusAIMS / COD / 02
Ages 8 to 12: making with AI, explained.
Open the syllabusAIMS / COD / 03
Pages a model can live on.
Open the syllabusAIMS / TEN / 01
Longer programs and real data.
Open the syllabusAIMS / TEN / 02
Teen-suitable tools, used with care.
Open the syllabusAIMS / TEN / 03
From one neuron to trained networks.
Open the syllabusHow lessons run
Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays.
Weekdays
After school and activities, in your US time zone.
Weekends
Weekend mornings, before the day fills up.
1-on-1
Paced for one child, at your convenience.
The same face every week.
Nothing that needs an adult account.
Discussion first, then a model they write.
What we covered and what to ask at home.
Live video, any state.
Fifteen to twenty in a batch, or one to one.
Student work
Projects from students who started at this age. The student labs page has more.

AI and ML
An AI nutrition coach that reads what you eat and works you toward a target.

AI and ML
A chatbot that answers mathematics and programming questions, built and deployed by a student.

AI and ML
An assistant that helps a young person recognise unsafe situations online.

Web app
A weather forecasting site with live conditions for any location.
Fees
School AI curricula are free and we encourage them. Our live classes are paid month to month in US dollars, with no term contract, so a family can stop after any month.
Free first class
USD 0
no card required
Group batch
USD 100
a month, billed in US dollars
One to one
USD 150
a month, billed in US dollars
What families say
Parents' words from Google, unchanged.
★★★★★
"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 parents
How models learn from examples, where they go wrong through bias and invented answers, how to check AI output, and a first small model in code. Using AI tools well is part of it, not all of it.
Some do, and free curricula make it easier. CodeAI publishes a free middle school course, AI Discoveries, with a unit on AI and machine learning, and Day of AI publishes a free grade-by-grade curriculum that covers grades 5 to 8. Ask your school which it uses.
With the right tools and supervision, yes. CodeAI says its Web Lab and AI tools are built for students ages 11 and up, and Machine Learning for Kids asks under-13s to use only a small language model with a trusted adult supervising.
Check the tool's own terms first, because age rules differ between services and change over time. For students under 13 we teach the ideas without adult chatbot accounts, using child-suitable tools and offline Python.
No. A good class starts with ideas and examples, then introduces a small amount of Python. Our first model on this page is about 20 lines.
Coding classes teach a programming language. AI classes teach how systems learn from data, using code as one tool. Doing both, in that order or side by side, works well.
It depends on the tool, so ask the school. CodeAI says AI Tutor interactions in AI Discoveries are automatically deleted after 90 days and never used to train third-party AI models.
An eighth grader who is 13 or older can usually join our teen Python course or AI tools course for ages 13 and up. Younger or newer students start with AI literacy for ages 8 to 14.
Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays.
The first class is free. After that, families pay monthly, and the fees section of this page shows the figure in US dollars.
Elsewhere on this site
More for US families with a middle schooler.
Our full set of US pages.
For younger children, ages 6 to 12.
The five abilities, age by age.
Builds that need no accounts.
The step up for eighth graders.
Every track, ages 6 to 14.
Ten AI projects for US middle schoolers, two in real Python, and how to take one to a science fair.
AI research programs for US high schoolers, from the 2026 to 2027 rules, with a fair-test experiment.
Four AI and Python certification exams for US high schoolers, compared from the issuers' own pages.
Start here
Tell us your child's grade and what they already do with AI. The teacher starts the first class from a question your child brings.
To see the approach first, read how we teach, or browse all our courses with full syllabuses.
WhatsApp us · +91 91233 66161 · connect@modernagecoders.com
Classes are live on video with teachers based in India, so families in any US state can join without travel. Message us on WhatsApp for a quick reply.
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.
Book a priority demo