Background: Community learning programmes may introduce generative AI without establishing whether participants can evaluate outputs or perform tasks independently. Aims: To develop a community teaching framework connecting AI use with verification and independent task performance. Review Method: A purposive reading of three official educational and accessibility resources informs conceptual design. No participant data, systematic review, or intervention results are claimed. Results and Conclusion: The proposed sequence links task definition, assisted production, explicit correction, and unaided transfer. Demonstrable judgement is distinguished from generated output volume. Effects depend on accessible reference materials, facilitator capability, and opportunities to practise. Contribution: The article offers operational definitions, a teaching example, assessment indicators, and testable propositions. The framework remains unvalidated.
Copyrights © 2026