How to teach AI literacy to older adults in British Columbia

What matters most

• Start with learner goals and confidence rather than AI features.
• Pair every useful task with a privacy and verification habit.
• Design for varying vision, hearing, dexterity, memory, language, device access, and digital experience.
• Treat an informed decision not to use AI as a valid learning outcome.
• Pilot with a community partner and improve from observed friction.

Begin with practical agency

A useful program helps an older adult decide whether an AI tool belongs in a task, complete a small task safely, and recognize when another source or a person is needed. It should not frame age as inability or assume that all learners share the same technology experience.

The Government of Canada’s seniors digital-inclusion report emphasizes tailored digital literacy, critical evaluation, privacy, security, accessibility, support, and community partnerships. Those themes translate directly into an AI literacy program: useful practice must arrive with safeguards and human support.

• Choose two or three low-consequence tasks such as drafting a general message, simplifying public information, creating a checklist, or translating non-sensitive text.
• Let learners observe before using a device and avoid mandatory account creation for essential concepts.
• Use fictional names, addresses, health details, payment information, and family scenarios during practice.
• Repeat the same safety routine across every session.

Teach four capabilities together

AI foundations, task skill, verification, and safety should not be separate optional topics. A learner who can produce an answer but cannot protect information or check an important claim has not completed the practical outcome.

A four-session sequence can introduce what AI is and is not, a safe question recipe, a stop-check-compare routine, and a pause-verify-tell-someone response to suspicious content. Each session should end with a printed take-home aid.

• Explain that generative AI creates plausible output and may be wrong even when it sounds confident.
• Show how to remove sensitive details and use general descriptions.
• Identify one material claim and compare it with an official or original source.
• Practise leaving an urgent interaction and reconnecting through a known trusted channel.

Design the room and interface for access

W3C guidance notes that accessible websites also address many needs associated with ageing, including changes in vision, dexterity, hearing, memory, and concentration. The course interface and the teaching room should therefore be treated as one learning system.

Use strong contrast, adjustable text, large controls, predictable navigation, captions, printed instructions, visible step numbers, and enough time to recover from an error. Avoid timed tasks, dense slides, tiny screenshots, and multi-step instructions delivered only once.

• Test the material with real learners using their own mix of devices and assistive tools.
• Have enough helpers for the group size and a private way to ask for support.
• Provide a non-digital backup for every essential exercise.
• Offer language support where the partner identifies a need.

Measure capability, safety, and partner value

Attendance and satisfaction do not prove that the program changed behaviour. Use a low-pressure baseline and completion check: can the learner remove sensitive information, complete one chosen task, verify a claim, and describe what to do with an urgent suspicious request?

For commercial validation, also ask whether the host can recruit the intended learners, whether delivery fits its operating model, what accessibility support was required, and whether the organization will sponsor, renew, or refer a second cohort. Report small pilot results as local observations, not conclusions about all older adults.

Continue with a practical resource