Universal Design for Learning, or UDL, is a framework for designing training that works for a range of learners from the start, instead of building one version and adding fixes for people it doesn't work for. CAST, the nonprofit that developed it, organizes UDL around three principles: engagement, representation, and action and expression.

Here's what each principle means, and what it looks like applied to a team learning AI at work.

The three principles

  • Engagement, the why: multiple ways to motivate learners and hold their attention, since what hooks one person can bore or worry another.
  • Representation, the what: multiple ways to present the same information, since some people take it in best by reading, some by watching, some by doing.
  • Action and expression, the how: multiple ways for learners to show what they've learned, instead of one fixed test format everyone has to fit.

CAST's own framing puts it simply: design for the variation that's already in the room, rather than treating that variation as something to accommodate later.

UDL applied to an AI training rollout

Say a company is training a mixed department, different roles, ages, and comfort levels with technology, on a new generative AI tool.

Engagement: some employees are curious and want to dig in immediately. Others are quietly worried the tool threatens their job. Offering both an opt-in advanced track and a low-pressure intro track respects both starting points instead of forcing everyone into one pace.

Representation: the same core lesson gets delivered three ways: a short written guide for people who'd rather read it once and reference it later, a five-minute recorded walkthrough for people who learn by watching, and a live hands-on session for people who need to try it with help nearby. Same content, three formats.

Action and expression: instead of one required quiz, people can show they've got it in different ways: record a short screen capture of themselves using the tool on a real task, walk a manager through it live, or submit a written before-and-after example. All three prove the same skill.

Why this matters specifically for AI training

Pew Research's June 2026 survey found about six in ten adults under 50 now use AI chatbots, compared with roughly four in ten adults 50 to 64 and fewer among adults 65 and older. Among workers who have used a chatbot, those under 50 were more likely than workers 50 and older to say it helped them work faster (44% versus 29%) and improved the quality of their work (31% versus 23%).

A single training format built for the most comfortable person in the room leaves a real, measured share of the workforce behind, and it isn't limited to age. Comfort with AI varies by role, past tech exposure, and plenty else. UDL designs the training with that range built into the plan from the start, before anyone has to struggle first.

How to use UDL if you're building AI training for a mixed team

Don't assume one starting point. Ask what your team already knows about AI before you build the session, the same first move the ADDIE model's Analysis step calls for.

Offer more than one way in. A written guide, a short video, and a live session covering the same material costs some extra prep time and reaches people a single format won't.

Let people prove the skill in more than one way. The point is whether they can use the tool on real work, not whether they can pass one specific test format.

That's the kind of training design behind the AI and digital literacy programs I build for teams and organizations. If you want help building AI training that works across a mixed team, the free 15-minute call is the place to start. You can also see the workshops and programs I run in the What I Offer section.

Sources

Notes marked as my process or professional opinion are exactly that, not external data.