What Are AI Lesson Planning Supports (and Why They Matter)

Lesson Enhancements are a set of supports that appear in every CARL lesson. They’re designed to surface things teachers deeply care about but that often get skipped when planning happens late at night or under time pressure.

Each CARL lesson includes prompts and suggestions across six areas:

1) UDL and Differentiation Pathways

Guiding question: How do we remove barriers and provide multiple access points for learning?

2) Equity and Inclusion Practices

Guiding question: Whose voices are centered? Whose experiences are acknowledged and honored?

3) Indigenous Voice Inclusion

Guiding question: How do we point teachers toward authentic Indigenous perspectives, not generic add-ons?

4) Social Emotional Learning (SEL) Focus

Guiding question: What emotional or relational support might students need during this activity?

5) Student Voice and Choice

Guiding question: Where can students make decisions that build real agency?

6) Trauma-Informed Safeguards

Guiding question: How do we create safety, predictability, and opt-out pathways for students who need them?

The key difference is timing and usability: CARL surfaces these supports inside the phase where you’ll actually use them, written in plain language you can drop straight into instruction. They’re options, not requirements.

Why the Enhancements Are Split into Two Types

One design choice that isn’t always obvious is that CARL separates enhancements into Teaching Enhancements and Learning Enhancements. That distinction matters—and it’s what makes CARL’s AI lesson planning supports different from generic add-ons.

• Teaching Enhancements appear after GRR, when you’re actively instructing, modeling, or guiding practice. They’re teacher-facing supports—things to think about while you’re teaching.
• Learning Enhancements show up during independent or collaborative work time (after Apply). They’re student-facing supports you can build directly into the learning experience.

Teachers told us they needed different kinds of support at different moments. So we designed the system to reflect that.

What That Looks Like in Practice

Here’s an example from a Grade 6 Social Studies lesson on political systems.

Teaching Enhancement (after GRR)

SEL Focus
Model “Expert Group Anxiety” during the I Do:

“Taking on the role of ‘teacher’ for my group feels like a lot of pressure. I’m going to look at my Evidence Template first and take a reset breath.”

This normalizes the discomfort of being responsible for teaching peers and shows students what self-regulation looks like before a challenging task.

Learning Enhancement (after Apply)

SEL Focus
Build in a Collaboration Thermometer midway through the island nation design:

“On a scale of 1-5, how well are you and your partner sharing the power of decision-making? If you’re at a 1, what’s one compromise you can make?”

This gives students a chance to assess their working relationship and practice negotiation before tensions escalate.

Same enhancement category. Different function. Different timing.

Why Specificity Matters

A lot of AI tools offer supports that sound reasonable but aren’t very usable. For example:

“Provide differentiation for diverse learners. Use sentence frames.”

That’s technically correct. It’s also not very helpful.

Here’s what CARL generates instead, from a Grade 12 Climate Science lesson:

UDL Pathways

Visual barrier
Feedback Loop Flowcharts with pre-cut arrows and labels, students physically map energy transfer steps on the board

Motor barrier
Digital whiteboard tools OR physical chart paper, student chooses the medium that lets them best visualize connections

Processing barrier
Component Word Bank with terms like “permafrost,” “methane,” “thermal energy” to reduce cognitive load while labeling complex feedback loops

That’s not a reminder to differentiate. That’s exactly how to adapt the activity.

Equity and Inclusion Practices

From a Grade 12 Climate Systems lesson:

When students model climate feedback loops, CARL suggests a built-in reflection prompt:

“If a community relying on permafrost for infrastructure stability reviewed your methane loop model, what human impacts would you want to add? Which geographic regions are most affected by your chosen feedback loop, and why is the impact unequal?”

This prompts students to consider Small Island Developing States facing rising sea levels from emissions they didn’t create, Arctic communities losing traditional food sources and infrastructure, and who has the least power to change the variables in the model yet faces the most immediate consequences.

Indigenous Voice Inclusion

From a Grade 10 English lesson on story structure:

When teaching the hero’s journey, CARL suggests exploring how story structure varies across cultures. Rather than treating Campbell’s monomyth as universal, invite students to learn from Indigenous authors and storytellers directly.

CARL points teachers to publicly shared, Indigenous-authored or Nation-approved resources like the FNESC Science First Peoples guides:

• The Back of the Turtle by Thomas King (Anishinaabe author)
• Monkey Beach by Eden Robinson (Haisla/Heiltsuk author)
• CBC’s Unreserved podcast episodes on storytelling traditions
• Your local First Nation’s website or cultural center for storytelling events and resources

That’s not “acknowledge Indigenous perspectives.” That’s pointing students directly to Indigenous voices and letting them draw their own conclusions from authentic sources.

Student Voice and Choice

From a Grade 6 Social Studies lesson on government systems:

After students complete their expert group research, CARL offers two levels of choice:

• Low-lift: “For your T-Chart comparison, choose which two government forms you found most interesting to compare, rather than being assigned the pairs.”
• Higher-lift: “What’s a rule in our school that feels like it was made by a ‘few’ instead of ‘many’? How would you change the decision-making process for that rule?”

Both honor student agency. One lets them choose what to analyze. The other lets them identify where the learning matters in their own life.

Safety Looks Different at Different Moments

Here’s an example from a Grade 2 picture book prediction lesson.

Trauma-Informed Safeguards (Teaching Enhancement)

Use a “Pass for Now” hand signal so students can opt to listen during group discussions without the pressure of being put on the spot when sharing predictions.

Trauma-Informed Safeguards (Learning Enhancement)

Make the Gallery Walk sharing step opt-in:

“You can post your illustration on the wall or submit it privately to me—you get equal credit either way.”

Both supports protect student safety. One reduces pressure during whole-class instruction. The other protects them when sharing creative work publicly.

Why We Care So Much About This

These enhancements reflect what I learned during my MEd in AI and Equitable Education, what emerged through a year of research with BC teachers, and what I believe makes good teaching actually good.

UDL, equity, trauma-informed practice, and student voice aren’t nice-to-haves. They’re foundational to classrooms where all students can learn.

But teachers don’t always have the capacity to plan for all of that, even when they care deeply about it. So we built CARL to surface these supports automatically.

Not as mandates. As options.

You might use all six enhancements in a lesson. You might use two. You might read them and decide today isn’t the day. That choice always belongs to the teacher.

How This Differs from Other AI Tools

A lot of AI tools surface supports as generic add-ons. CARL treats them as structural, phase-aware, and activity-specific.

They’re generated for every lesson, tailored to the specific activities, and split by instructional phase so they show up when they’re actually useful. They’re also written in plain language you can use immediately.

From a Grade 12 Climate lesson:

Trauma-Informed Safeguards

If modeling tipping points feels distressing, students can switch to the “Stabilization” tutorial mode showing negative feedback loops that balance systems, focusing on hopeful data instead.

That’s not “consider trauma-informed practice.” That’s exactly how to provide an exit path.