Page 278 - CyberSecurityEssentialsEnglish

▣ A one-task policy template
Permitted: using the tool to understand instructions, generate revision questions,
improve clarity of language after writing a personal draft, or propose examples that you
then verify.
Conditional: using the tool to organize the argument or compare alternatives, provided
you disclose it and state what was changed or rejected.

Prohibited: generating the complete final answer, fabricating sources, entering sensitive
data, or presenting the tool’s output as unreviewed personal work.
Disclosure required: the name of the tool, the purpose, where it was used, its effect on
the final version, and what the learner verified themselves.

6.8 Accessibility, Inclusive Learning, and Different Needs
◆ Key idea: inclusion removes barriers; it does not multiply versions.

The inclusiveness of teaching is not measured by the number of versions AI produces, but
by its ability to remove barriers and enable participation and expression, while preserving
a meaningful educational objective, the learner’s dignity, and their choice.

AI may open important opportunities for learners with disabilities, for those learning in a
second language, and for anyone who needs organization, simplification, or conversion
between media. It can create a first description of an image, summarize a text in clear

language, turn speech into text, propose several examples, or reorganize a task. But an
automated output is not automatically accessible, and it may contain an incomplete
description, confusing language, or an order unsuited to a screen reader.

Universal design for learning offers a framework for building experiences that reduce

barriers from the outset. Current guidance focuses on multiple means of engagement,
representation, and action and expression, and aims to strengthen learner agency and the
capacity to choose and reflect [6][19]. This does not mean producing a separate lesson for

each person; it means offering varied ways to access, take part, and demonstrate learning.
Engagement concerns motivation, identity, belonging, and choice. AI may help connect an
activity to different interests or propose roles within a group, but personalization based on

stereotypes or sensitive inferences must be avoided. The learner must remain able to choose
a suitable path rather than be confined to a profile that predicts what suits them.

Representation concerns how information is presented. Text, audio, diagrams, tables, and

examples can be offered, with vocabulary and symbols explained. The visual form must not
be mere decoration, and the meaning must not disappear when content is converted. An
equation needs a description of its structure, and a diagram needs alternative text that

states the essential relationship, not only the colors.

Action and expression concern how the learner shows what they know. The objective may
be demonstrable through writing, a presentation, a spoken explanation, a performance, or

277