How can teachers help middle-years students verify AI output and explain the choices they make about using it?
Design classroom tasks that help middle-years students verify AI output, recognise its limitations and explain how its use changes their work.
Many students can produce something with generative AI. Far fewer can explain whether the output is reliable, how it may be biased or what using it has changed about their own thinking.
This course helps teachers design practical classroom tasks in which students test AI output, make deliberate choices and account for the part AI has played in their work.
Give students practical ways to question AI output and explain the choices they make.
Session 1
Participants test how changes to a prompt, its context and its constraints alter an AI response. They document the changes and examine how a system can produce an answer with high confidence while still getting it wrong.
The work is based on practical comparison: what changed, why it may have changed, and what that means for a student using the result.
Session 2
Participants work with AI-generated material containing correct and incorrect claims. They distinguish factual claims from opinions and value judgements, identify factual errors and weak reasoning, then ask why a system can present unreliable information so fluently.
They use Faktabaari’s three-question routine to check unfamiliar material: who is behind the information, what is the claim based on, and what do other sources say? The checking happens laterally, by looking beyond the original page or answer.
Session 3
Participants use a simple framework for making AI use visible: what was asked, what was returned, what was changed and why. They apply it to an existing classroom task, then consider whether the tool supported pupils’ thinking or replaced it.
The session focuses on transparent explanation rather than AI detection. Pupils need language for describing their own contribution, the machine’s contribution and the decisions they made.
Follow-up
A 60-minute online review of one classroom trial. Participants bring a pupil explanation, a prompt-and-output comparison or a source-checking exercise, then identify one adjustment for the next version.
Starts 17 February 2027
3 sessions: 17 and 24 February, 3 March 2027
Online
Starts 22 June 2027
3 sessions: 22 and 29 June, 6 July 2027
Online
Starts 8 September 2027
3 sessions: 8, 15 and 22 September 2027
Online
Starts 10 November 2027
3 sessions: 10, 17 and 24 November 2027
Online
Registering interest is free. A course run is confirmed once at least six participants have registered their interest. Until then, neither a place nor a course run is guaranteed. Booking terms →
Run the course with your own staff, on your own dates.
Session 1
Participants test how changes to a prompt, its context and its constraints alter an AI response. They document the changes and examine how a system can produce an answer with high confidence while still getting it wrong. The work is based on practical comparison: what changed, why it may have changed, and what that means for a student using the result.
Session 2
Participants work with AI-generated material containing correct and incorrect claims. They distinguish factual claims from opinions and value judgements, identify factual errors and weak reasoning, then ask why a system can present unreliable information so fluently. They use Faktabaari’s three-question routine to check unfamiliar material: who is behind the information, what is the claim based on, and what do other sources say? The checking happens laterally, by looking beyond the original page or answer.
Session 3
Participants use a simple framework for making AI use visible: what was asked, what was returned, what was changed and why. They apply it to an existing classroom task, then consider whether the tool supported pupils’ thinking or replaced it. The session focuses on transparent explanation rather than AI detection. Pupils need language for describing their own contribution, the machine’s contribution and the decisions they made.
Follow-up
A 60-minute online review of one classroom trial. Participants bring a pupil explanation, a prompt-and-output comparison or a source-checking exercise, then identify one adjustment for the next version.
Every course we run is built on what research says makes professional learning change practice. Here is what that looks like in this one.
Hands-on investigation of shared AI examples: how generative and predictive systems work, and where bias and privacy enter.
You choose one real curriculum task to redesign so that students can explain and account for their AI decisions.
Fixed small groups work through classroom cases, give feedback on each other's redesigns and rehearse the approach.
You test part of your design with students, then use a live online follow-up session to review the evidence and refine it.
Three live online sessions, one week apart, move from understanding and interrogation to practical curriculum design. Teachers work with shared examples, apply the ideas to their own learners and receive structured peer feedback. An optional group follow-up supports refinement after the course.
The course family is informed by the OECD and European Commission AI Literacy Framework, including the Engage, Create, Manage and Shape dimensions, alongside Nordic AI literacy work including Faktabaari EDU. The Nordic Learning Company translates these reference points into age-appropriate classroom learning and professional judgement about when AI supports learning and when it replaces intellectual work students need to do themselves.
EUR 350 per participant, excl. VAT
Course materials, facilitation, a certificate of attendance and an online follow-up session.
Nick Praulins is co-founder of The Nordic Learning Company, where he leads our work with schools and educators, shaping the learning philosophy, pedagogy and programme design behind its education practice.
View profile →Marius Ketels is co-founder of The Nordic Learning Company, where he leads much of our work with companies and organisations, designing and facilitating programmes around leadership, teams and human performance.
View profile →You can move to another date or send a colleague at no cost. If you cancel more than four weeks before the start, we keep 30% of the fee; inside four weeks, we keep the full fee. If we cancel, you choose a full refund or another date.
Pairs well with
Practical AI fluency and the professional judgement to decide where AI belongs in teaching and learning.
Help children become curious, questioning and increasingly thoughtful about the AI already around them.
Build an app. Learn to teach app building.
EUR 350 per participant, excl. VAT
Starts 17 February 2027
