Learn to Think, Not Copy AI: Coding Classes in Germany

In short

Learning to think, not copy AI means a learner plans a solution, predicts what it will do, tests it and can explain it, and treats any AI answer as a claim to check. Modern Age Coders teaches coding and problem solving this way for children and teenagers in Germany, live online in English, in a group or one to one, starting with the Python and AI for kids course.

Courses that build the habit

Three courses where thinking comes first

Each starts with the learner's own reasoning and adds tools only when the reasoning is there.

LTO / 01

Python and AI for kids

Ages 8 to 14. Real Python, small programs that learn from data, and the question "how would we know if it is wrong?" in every class.

Open the syllabus →

LTO / 02

Problem solving and computational thinking

Puzzles and patterns that train breaking a problem down before writing a line of code.

Open the syllabus →

LTO / 03

AI literacy for kids

What AI is, where it fails, and how to check an answer, without needing an adult-only chatbot account.

Open the syllabus →

What German schools say

What do German education ministers say about AI in learning?

On 10 October 2024 the education ministers of the Länder adopted a recommendation on how schools should handle artificial intelligence. It is a careful document, and on one point it is very clear: the aim is pupils who can judge AI, not pupils who only use it.

No ban, but judgement

The recommendation states that a general ban on using AI for homework is neither useful nor desirable nor enforceable, given the world pupils live in. Instead, teachers and pupils are to agree how AI may be used and how AI-generated work is marked as such.

In learning about AI, the KMK says, what matters is being able to assess the quality of AI results and their effects, including from an ethical point of view.

Exams that show your own work

Where a pupil uses AI to make something and then presents it, the recommendation asks that the defence also assess how skilfully the pupil worked with the AI and how well they can reflect on the results.

Exam formats that cannot legally guarantee the independent work a task demands are to be abolished or fundamentally reworked. Marking, the KMK adds, is and remains a task for teachers.

The document also records a disagreement worth knowing. The KMK's own scientific commission, the SWK, published a paper on large language models on 17 January 2024 and recommended largely doing without AI language models in primary school and the first years of lower secondary. The ministers call that recommendation one that needs discussion, given the potential of adaptive learning tools. In other words, the experts and the ministers agree that judgement matters and disagree on how early tools should arrive.

Source: KMK, Handlungsempfehlung für die Bildungsverwaltung zum Umgang mit Künstlicher Intelligenz in schulischen Bildungsprozessen, Beschluss der Bildungsministerkonferenz vom 10.10.2024, revised edition March 2025, read 3 November 2026. Our English summaries; the German text is the authority.

The 2026 framework

What does the KMK's 2026 AI framework expect pupils to do?

On 16 October 2026 the KMK adopted an update of its competence framework Bildung in der digitalen Welt for the age of AI. The six areas from 2016 stay the same; AI is added inside them, under four headings: dealing with AI consciously, using it creatively, using it purposefully and helping to shape it. Five of the new lines describe exactly what we mean by thinking, not copying.

Five AI-related competences in the KMK framework of 16 October 2026, and how they look in a coding class
No.	The KMK's words	In plain English	In our classes
1.2.1	reflektierte Einordnung KI-generierter Ergebnisse sowie möglicher Verzerrungen	Judge AI results and spot bias	Every answer, ours included, gets tested before it is believed
1.2.2	Informationsquellen auch im Abgleich mit Primärquellen analysieren	Check against the original source	We compare a claim with the law, the data or the documentation itself
2.2.2	transparente Kennzeichnung KI-generierter und -überarbeiteter Beiträge	Say openly what AI made	Learners mark any code they did not write themselves
3.2.2	KI-unterstützte Ergebnisse prüfen, anpassen und begründet in eigene Produkte einbinden	Check, adapt, and justify what you use	A line of code stays only if the learner can explain it
5.5.1	zentraler Konzepte maschinellen Lernens sowie deren Grenzen	Know how machine learning works and where it stops	Learners train small models and find where they fail

Source: KMK-Kompetenzrahmen "Bildung in der digitalen Welt"; Aktualisierung im Kontext Künstlicher Intelligenz (2026), Beschluss der Bildungsministerkonferenz vom 16.10.2026, read 3 November 2026. The framework says it draws on the OECD and European Commission AI Literacy Framework published in 2026. The right-hand column is ours.

One more line in the framework, 5.5.3, asks pupils to plan and use a structured, algorithmic sequence to solve a problem. That is programming in all but name, and it comes before any of the AI lines in practice: a learner who cannot plan steps cannot check whether an AI's steps are right.

An exercise in checking

Can a confident answer be right this year and wrong last year?

Our own exercise. Ask "when does summer time start in Germany?" and you will often hear a tidy rule: the fourth Sunday in March. It sounds official and it gives the right date for 2027. The German summer time ordinance (Sommerzeitverordnung, section 2) says something different: summer time begins on the last Sunday in March at 2:00. We wrote both rules in Python and compared them over twenty-one years.

A confident rule against the lawPython

from datetime import date, timedelta

def confident_answer(year):
    # "Summer time starts on the fourth Sunday in March."
    day = date(year, 3, 1)
    while day.weekday() != 6:          # 6 means Sunday
        day += timedelta(days=1)
    return day + timedelta(weeks=3)

def the_law(year):
    # SoZV section 2: "am letzten Sonntag im März"
    day = date(year, 3, 31)
    while day.weekday() != 6:
        day -= timedelta(days=1)
    return day

print("2027:", confident_answer(2027), the_law(2027))
wrong = [y for y in range(2020, 2041) if confident_answer(y) != the_law(y)]
print("Years 2020 to 2040 where the answer is wrong:", wrong)

Run with Python 3 on 3 November 2026, it prints 2027: 2027-03-28 2027-03-28 and Years 2020 to 2040 where the answer is wrong: [2020, 2024, 2025, 2026, 2030, 2031, 2036, 2037].

Why it fooled us in 2027

March always has 31 days, so it has either four or five Sundays. When it has four, the fourth Sunday is the last one and the tidy rule works, as it does on 28 March 2027. When March has five Sundays, the rule picks the wrong one, a week too early.

In 2026, March began on a Sunday, so it had five: the rule says 22 March, the law says 29 March 2026. Over 2020 to 2040 the tidy rule fails in eight years of twenty-one.

What the exercise teaches

Checking the answer once, for the year you happen to care about, would have confirmed a wrong rule. Checking it against the definition, over many cases, exposed it in seconds.

That is the habit the KMK framework describes as checking against primary sources. The source here is a German ordinance anyone can read on gesetze-im-internet.de; the check is a few lines of Python.

Ages 8 to 10, on paper. Take a calendar of March 2026. Circle every Sunday. Which is the fourth? Which is the last? Are they the same?
Ages 10 to 13, in Python. Run the program, then change the range to 2000 to 2100 and count the wrong years. Is it roughly the same share?
Ages 13 and up. The ordinance says summer time ends on the last Sunday in October. Write end_of_summer_time(year) and test it the same way: against the definition, over many years, before you trust it.

Competitions

Are AI tools allowed in German maths and coding competitions?

It depends on the competition, and the rules are getting sharper.

A flat no in the Bundeswettbewerb Mathematik

The Bundeswettbewerb Mathematik, a national maths competition for pupils on the way to the Abitur, lists using artificial intelligence to solve the problems, or even to check your own solution, among the things participants must not do.

A pupil who has only ever checked work with a chatbot will find that rule hard. One who checks by testing cases, as in the exercise above, is already used to it.

Where building AI is the task

The opposite case is the Bundeswettbewerb Künstliche Intelligenz (BWKI), where pupils must build a project with machine-learning methods. Our page on AI and machine learning in Germany covers it.

Both rules point the same way: a competition wants to see the pupil's own thinking, whether the subject is proofs or models.

Our method

How do our classes teach children to think instead of copying?

The owner of Modern Age Coders promises families the same teaching anywhere in the world: we teach learners how to think and solve problems, not how to copy answers or AI output. In a live class that promise becomes five routines.

Say it before you run it. Before any program runs, the learner predicts the output out loud. A wrong prediction is the most useful moment in the class.
Plan in words first. The learner writes the steps in plain English, then turns them into code. The KMK's framework calls this a structured, algorithmic sequence.
Test the edges. The first day of a month, an empty list, a five-Sunday March. Learners choose the awkward cases themselves.
Explain every line. Code the learner cannot explain is deleted and written again, however it got there.
Check against the source. A rule is compared with the law, the dataset or the documentation it claims to describe.

English only, live

Every class runs live and in English. We do not teach in German; the mentor names German school terms where they help.

Pace and accent matched

Mentors wait for the learner's own idea instead of supplying the answer, and adjust to how the learner speaks.

The same mentor

One person who notices when a learner starts guessing and brings them back to reasoning.

Tools and ages

Which AI tools are suitable for children in Germany?

We do not recommend any AI chatbot to children, and our classes for under-18s do not need one.

Read the provider's own age rules

Providers set their own minimum ages. Anthropic's consumer terms for Claude require users to be at least 18. OpenAI's terms for ChatGPT set 13 as the floor and require a parent's or guardian's permission for anyone under 18. Our AI tools age guide quotes these terms for many tools.

The KMK recommendation also records the Länder's aim of giving schools data-protection-compliant, free access to language models, which would answer the question differently inside school.

What children use with us instead

Children write Python, train small models on data they can see, and check results against sources. When AI comes into a lesson, it is something the learner builds and tests, not an adult account they log into.

Older teenagers learn how language models work and where they fail, so that when they do use one at sixteen or eighteen, they use it as the framework intends: checked, adapted and declared.

A path

From first program to checking any answer

Our own stages for the thinking habit, from age eight upwards.

Thinking milestones (ours)
Stage	Typical age	The learner can
Predict	8 to 9	Say what a short program will do before running it
Plan	9 to 11	Write the steps of a solution in words, then in code
Test	11 to 13	Choose awkward cases and check a rule against them
Explain	12 to 15	Justify every line of a program to someone else
Judge	14 and up	Check an AI answer against a source and say what is wrong with it

For parents

Ask your child to predict before they press run. It is the simplest habit to support at home.

If homework involved AI, ask which part your child could explain without it.

For learners

When an answer looks perfect, try the strangest case you can think of.

Write down where an idea came from. Your school may ask, and the KMK expects it.

More courses

Nine more courses that put reasoning first

Live in English with a regular mentor; choose a batch or private lessons.

I

Young thinkers

Ages 6 to 11

LTO / YOUNG / 01

Scratch

Planning and debugging with blocks.

Open the syllabus

LTO / YOUNG / 02

Reasoning for young thinkers

Logic and patterns before code.

Open the syllabus

LTO / YOUNG / 03

Mathematik-Olympiade and Känguru

Competition problems, our own, no AI.

Open the syllabus

II

Teen programmers

Ages 11 to 16

LTO / TEEN / 01

Python for teens

Programs they can explain line by line.

Open the syllabus

LTO / TEEN / 02

Maths through coding

Test maths ideas in code.

Open the syllabus

LTO / TEEN / 03

Informatik Companion

Alongside school Informatik, Klasse 5 to 10.

Open the syllabus

III

Building, not prompting

Ages 13 to 18

LTO / BUILD / 01

AI and machine learning

Train models and find where they fail.

Open the syllabus

LTO / BUILD / 02

Algorithms and data structures

The reasoning behind fast programs.

Open the syllabus

LTO / BUILD / 03

Biber, BWINF and BWKI

German competitions, our own practice problems.

Open the syllabus

How classes run

How a live thinking-first class works

Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays. Each class follows the same rhythm: a problem, a prediction, the learner's own attempt, then a test.

After school

For example 14:15 German time, which is 18:45 in India in winter.

Later

For example 16:15 German time, 19:45 in India from 28 March 2027.

1-on-1

Saturday at 11:00, or a time your family picks.

Problem first

Every class opens with a question, not a tutorial.

Predict, then run

Learners commit to an answer before the computer gives one.

Edge cases

The awkward input is part of every task.

Own words

Learners explain their code back to the mentor.

Parents informed

Short notes on what was reasoned out, not just finished.

Batch or private

Fifteen to twenty learners, or private lessons.

Student work

What our students have made themselves

These projects were planned, built and explained by our students. A selection is below; the student labs page has more.

AI and ML

NutriLife

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

by Bhavya · Open it

AI and ML

Misti

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

by Harshit · Open it

AI and ML

GuardianX

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

by Vivaan · Open it

Web app

SkyCast

A weather forecasting site with live conditions for any location.

by Krish

Fees

Fees for families in Germany

You pay per month for eight live classes. Prices here are euro amounts based on the European Central Bank reference rate; the actual charge is in US dollars and your bank converts it, so a small difference on the statement is normal.

Free first class

Free

no card required

A free first class
A short problem to see how your child reasons
No card needed

Book it

Group batch

€90

a month, eight live classes

Fifteen to twenty learners at one level
One regular mentor
Two classes a week
Monthly, cancel any month

Start here

One to one

€134

a month, eight live classes

Your child and one mentor
Paced to how your child thinks
Times your family chooses

Enquire

What families say

Rated 4.9 across 547 Google reviews

Google reviews in the words of the parents and students who wrote them.

★★★★★

"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

Parents ask

Thinking, not copying AI: common questions

What does "learn to think, not copy AI" mean in practice?

The learner plans a solution, predicts what it will do, tests it on awkward cases and explains every line. Any answer from a tool, or from the mentor, is treated as a claim to check, not a result to copy.

Is AI banned in German schools?

No. The KMK recommendation of 10 October 2024 says a general ban on AI for homework is neither useful nor enforceable, and asks schools to agree how AI may be used and how AI-generated work is labelled.

What does the KMK expect pupils to be able to do with AI?

Its framework of 16 October 2026 adds AI to the six areas of Bildung in der digitalen Welt: for example, judging AI results and possible bias, checking against primary sources, labelling AI-generated work and checking and adapting AI results before using them.

Should my child use ChatGPT or Claude for coding homework?

We do not recommend any chatbot to children. Claude's terms require users to be 18, and ChatGPT's terms require a parent's permission for anyone under 18. Our classes for children need neither.

Are AI tools allowed in the Bundeswettbewerb Mathematik?

No. The competition forbids using artificial intelligence to solve its problems or to check your own solutions.

Will my child fall behind without AI tools?

Our view is the reverse: a learner who can plan, test and explain will use any tool better later, and the KMK framework asks for exactly those skills. Children with us still learn how AI works by building small models.

Are the classes in German?

No, the teaching is in English only and always live. Mentors work from India and adapt to each child's pace and accent. For school-related topics they also give the German term.

When are the classes?

Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays. Our mentors' day in India runs 4 hours 30 minutes ahead of Germany in winter and 3 hours 30 minutes in summer.

Are you connected to the KMK or a school?

No. Modern Age Coders is not connected to the KMK, the SWK, any ministry or competition. We teach alongside school; school marks come only from the school.

How much does it cost?

A monthly fee, displayed here in euro at the European Central Bank reference rate and charged in US dollars, which your bank converts. The first class is free.

Elsewhere on this site

Related pages

More on thinking, coding and AI for learners in Germany.

AI and machine learning in Germany

Building and testing models, and the BWKI.

AI tools age guide

Minimum ages from the providers' own terms.

How to think like a programmer

The habits behind good code.

Coding classes in Germany

All our pages for families in Germany.

Informatik-Nachhilfe online

School Informatik help, page in German.

Python and AI for kids syllabus

Every topic in the course.

Start here

Book a free first class

The first class is one small problem. The mentor watches how your child approaches it, without giving the answer, and then suggests a level and a plan.

Read first, if you like: the Python and AI for kids syllabus and how we teach.

WhatsApp us · +91 91233 66161 · connect@modernagecoders.com

Classes take place live online in English, taught by mentors in India. The fastest way to reach us is WhatsApp.

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.

Book a priority demo

Three courses where thinking comes first

Python and AI for kids

Problem solving and computational thinking

AI literacy for kids

What do German education ministers say about AI in learning?

No ban, but judgement

Exams that show your own work

What does the KMK's 2026 AI framework expect pupils to do?

Can a confident answer be right this year and wrong last year?

Why it fooled us in 2027

What the exercise teaches

Are AI tools allowed in German maths and coding competitions?

A flat no in the Bundeswettbewerb Mathematik

Where building AI is the task

How do our classes teach children to think instead of copying?

English only, live

Pace and accent matched

The same mentor

Which AI tools are suitable for children in Germany?

Read the provider's own age rules

What children use with us instead

From first program to checking any answer

For parents

For learners

Nine more courses that put reasoning first

Young thinkers

Scratch

Reasoning for young thinkers

Mathematik-Olympiade and Känguru

Teen programmers

Python for teens

Maths through coding

Informatik Companion

Building, not prompting

AI and machine learning

Algorithms and data structures

Biber, BWINF and BWKI

How a live thinking-first class works

Problem first

Predict, then run

Edge cases

Own words

Parents informed

Batch or private

What our students have made themselves

NutriLife

Misti

GuardianX

SkyCast

Fees for families in Germany

Rated 4.9 across 547 Google reviews

Thinking, not copying AI: common questions

What does "learn to think, not copy AI" mean in practice?

Is AI banned in German schools?

What does the KMK expect pupils to be able to do with AI?

Should my child use ChatGPT or Claude for coding homework?

Are AI tools allowed in the Bundeswettbewerb Mathematik?

Will my child fall behind without AI tools?

Are the classes in German?

When are the classes?

Are you connected to the KMK or a school?

How much does it cost?

Related pages

AI and machine learning in Germany

AI tools age guide

How to think like a programmer

Coding classes in Germany

Informatik-Nachhilfe online

Python and AI for kids syllabus

Book a free first class

Want a class sooner? Book a priority demo

Keep exploring Modern Age Coders

Related courses and guides

Learn more

Free resources

From the blog

Start here