Teaching AI Across Business, English and Art
Teaching AI Across Business, English and Art
AI arrived in school faster than any scheme of work. Business classes ask what happens to jobs, English classes ask who wrote this, art and music classes have argued about borrowing for a century. This page connects those units into one sequence for Grades 9 to 12.
See the unit →Resources that fit
Units and bundles for this topic
Start with the unit that matches your next teaching block; the bundle is there if you need the whole strand. Tap any cover for the full contents, preview and price.
Digitalisation and AI in the World of Work
Media and Digitalisation – Chances and Risks
Digital Image Editing Activities | Pixels, Photo Editing & Media Ethics | Middle School Art Unit | Grades 7–9
Music About Music | Quotation, Homage, Parody & Sampling | Complete High School Music Unit
Data protection and the GDPR at the workplace
The teaching problem
Teaching AI When Students Already Use It
The usual difficulty with a new topic is that nobody in the room knows anything. Here the opposite holds: most of the class has used these systems more recently than you have, and they have absorbed a folk model along the way. That model says the tool looks things up, so a confident answer is a checked answer. Once a student believes that, a fabricated citation is not a warning sign, it is just an answer they did not verify. The second problem is pace. Anything built around a named product is stale within a year, which is why the durable content sits elsewhere: how a predictive model is trained, what consent to data use means legally, and the long argument about quotation and authorship that art and music have been running since collage.
A sequence that works
Five Lessons That Outlast the Hype
Nothing here depends on a particular product or a login. The sequence works from how prediction from training data behaves, out to the workplace, the law, and the older question of what counts as your own work.
- What the model is doingStudents run a paper prediction game, guessing the next word from a short corpus, then explain why a system built this way produces fluent text and unreliable facts.
- Tasks, not jobs, changeA single occupation is broken into its actual tasks and each one is sorted by how automatable it looks. The class then argues about the three or four hardest to place.
- Whose data went inWorkplace monitoring and consent are examined through the data protection unit, with students deciding which of six scenarios a company could lawfully defend and on what basis.
- Sampling, quotation and generated imagesMusical borrowing and photomontage give the class a century of precedent. Students then place four cases on a line running from homage to theft and justify each placement.
- Writing a position with evidenceThe unit closes with an argumentative piece on one workplace or classroom rule, requiring a counterargument and at least one source the student has actually opened.
Where it goes wrong
Assessment When the Tool Is Available
Set a handwritten baseline in the first week so you know each student's unaided voice; without it, every later suspicion is a guess. Assess the process rather than the polish: a research log, a marked-up draft, a two-minute oral defense of one paragraph. On content, the errors are predictable. Students report that the system searched for an answer, cite sources that do not exist, and treat fluent prose as evidence of accuracy. Make verification a graded step, with a rule that any cited source must be linked and quoted. Detection software is not reliable enough to carry a grade dispute, so do not build your policy on it.
What's in the download
Inside the files
Editable Word and PowerPoint plus print-ready PDFs, with answer keys throughout.
- AI and world of work unit
- Data protection case scenarios
- Image editing and media ethics tasks
- Sampling and quotation listening sheet
- Argumentative writing frame
- Assessment and class policy templates
Good to know
Frequently asked questions
Will this material be out of date next year?
Parts of any AI resource date quickly, which is why the lessons avoid naming products and screenshots of interfaces. What the sequence teaches is how prediction from training data behaves, what the law says about consent and workplace monitoring, and how authorship arguments have run since collage and sampling. Those hold. The files are editable Word and PowerPoint, so when an example ages you can replace it in a few minutes.
Do students need accounts or access to AI tools?
No. Every task in the sequence works on paper, including the prediction game in lesson one, which is deliberately done by hand so the mechanism stays visible. If your students do have access and school policy permits it, the verification task in lesson five becomes sharper when they check real output against real sources. It is an option rather than a requirement, and nothing breaks without it.
Can I teach this without a technical background?
Yes. The one technical idea, that a model predicts likely continuations rather than retrieving stored facts, is carried by the paper activity and explained in the notes. Nothing asks you to discuss architecture or training methods. The heavier lifting is in the areas most teachers already handle: reading a legal scenario, weighing a claim, and marking an argument that has to survive a counterargument.
A Sequence That Ages Well
Teach how the system works, who owns the data and what counts as your own work, and next year's product changes nothing.
Browse the full collection →