AI literacy activities for Science lessons

• Classification connects this Science lesson to AI: both people and computer systems group things by their features.
• Pupils test how sorting rules can fail on unusual animals, then discuss why an AI classifier also needs checking.

• Classifier: a system that assigns items to groups.
• Feature: a detail used to tell one item from another.
• Training data: examples a computer system learns from.
• Bias: a pattern of unfair or inaccurate results when examples are limited.

• UNESCO: understand and question how AI systems make decisions.
• OECD/EU AI Literacy: recognise how data and design affect outputs.
• PISA 2029: evaluate claims and evidence about AI results.
• Anthropic AI Fluency: check an AI output before relying on it.

Illustrative links. A generated guide gives the detailed mapping for your lesson and region.

• An AI system can use examples to classify or predict, just as pupils use observations to explain what happens in a leaf. Scientific evidence and human judgement still check the output.
• Pupils test how missing or unbalanced examples change a prediction, and why a pattern match is not a scientific explanation.

• Training data: examples given to an AI system so it can find patterns.
• Classifier: an AI system that sorts an input into a named group.
• Model: the pattern system produced after training examples have been used.
• Prediction: an output estimating which group or result is most likely.

• UNESCO: compare pattern matching with the evidence needed to explain a result.
• OECD/EU AI Literacy: accept, revise or reject an AI output by checking it against evidence.
• PISA 2029: identify how missing or unbalanced data can bias a prediction.
• UNESCO AI system design: design a system that keeps a human check on its output.

Illustrative links. A generated guide gives the detailed mapping for your lesson and region.