This is the second post in our series exploring the four pillars from Guidance for the Use of AI in the K12 Classroom, a report authored by SREB Project-Based Learning Director Leslie Eaves for the SREB Commission on AI in education.

(If you missed it, make sure to read the first post in our series, watch the video or listen to the podcast episode.)

In this post—and the accompanying podcast and YouTube video—I’m joined again by Leslie Eaves to talk about creating cognitively demanding tasks.

 

Using AI to Create Cognitively Demanding Tasks

The report encourages educators to use AI tools to design assignments that require students to engage in deeper thinking, creativity, problem-solving and innovative reasoning.

These are not simple worksheets, fill-in-the-blanks or regurgitation of facts. Instead, they are thoughtfully crafted learning experiences that guide students to analyze, evaluate, create and apply knowledge in meaningful ways.

As Leslie explains in the podcast, this is about designing a learning process that students “can’t opt out of”—work they can’t just feed into a chatbot and get a polished answer in return. The goal is to create tasks that require student thought, judgment and synthesis—tasks where AI can support the process but not replace the learning.

Let’s break down what cognitively demanding tasks actually look like, how you can use AI to plan them and where to start—even if you’ve never used AI tools in your classroom before.

 

What Is a Cognitively Demanding Task?

Let’s start by defining the term: assignments that truly engage students in the kind of creative, analytical and extended thinking they’ll need in life and in the workplace.

This kind of thinking is at the heart of project-based learning — but it doesn’t require a full project-based learning unit to get started. Any task that pushes students into the “analyze,” “evaluate,” or “create” zones of Bloom’s Taxonomy—and requires extended reasoning over time—qualifies.

One simple example: Instead of having students memorize state capitals (something AI can do instantly), ask them to analyze why certain cities were chosen as capitals. Even better: Have them propose a new capital based on historical, geographic and political criteria.

 

How Can AI Help?

AI can be a useful tool for designing tasks that will “really engage their brains,” Leslie said. “AI can help you create an outline, but it’s not going to know your students, it’s not going to know their passions and curiosities.”

In other words, AI can help you brainstorm good ideas and plans, but those lessons ultimately needed to be guided by you, the one who knows what will work for your classroom. AI is just your assistant.

Another practical example Leslie relates is a third-grade community project teachers used AI to revise. They entered their standards into an AI tool and asked it to generate a problem-based unit. The output gave them a helpful starting point — with phases and prompts they could adapt and refine.

Even for smaller assignments, here are some practical ways you can start using AI today:

• Draft task outlines that align to your standards
• Summarize research materials for lower reading levels
• Find real-world examples that resonate with your students, especially for topics you know nothing about—like sports or anime!
• Scaffold complex tasks with step-by-step tools or graphic organizers

A Practical Example for Teachers

When it comes to any new tool or idea, getting started can be the hardest part. This is why Leslie recommends starting with your standards. When standards use verbs like analyze, design or evaluate, these often signal higher-order thinking goals.

1. Write a prompt.
   For example, “Create a project-based learning assignment aligned to [standard]. The task should require students to engage in extended reasoning and culminate in a real-world product or presentation.”
2. Review the output critically.
   Use it as a draft — not a final plan. Adjust the task for your students, your timeline.
3. Chunk the assignment into steps.
   Add activities that guide students through research, brainstorming and feedback—so they don’t just jump to using AI themselves.
4. Build in support tools.
   Need a custom reading at an eighth-grade level? A new rubric? A visual brainstorming sheet? AI can help with those, too.

“AI can be a great thought partner,” Leslie said.

(For more on AI prompts, listen to our first-ever podcast episode: “Easy Lesson Planning With Spice-Y AI Prompts.”)

Keep These Cautions in Mind

Leslie also shared a few cautions:

• Don’t over-rely on AI. Your classroom shouldn’t fall apart if the internet goes down. Offline thinking and discussion still matter.
• Be mindful of access. Not every student or school has the same AI tools available. Don’t assume.
• Start small. One well-designed, high-level task is better than five shallow ones.

Most importantly, AI is a part of your journey towards creating cognitively demanding tasks for your students, “but it cannot be the end-all and be-all of that journey.”

Finally, Leslie emphasized one major caution, even going so far as to say this isn’t a caution so much as a command: Do NOT put in any identifying information about a student!

 

Try This: One Small Step to Start

Leslie suggested that you try picking one set of standards you’ll be teaching this fall.

Use an AI tool—like ChatGPT, Microsoft Copilot or MagicSchool—to draft a task outline. Then refine it into something that would get your students thinking deeply.

 

Want More?

This episode is part of a full series on how AI can support teachers, not overwhelm them. If you prefer listening or watching:

• Listen to the podcast
• Watch the video
• Download the full free report that inspired this series
• Download our AI Tool Procurement, Implementation and Evaluation Checklist

Finally, keep up with the latest from our Commission on AI in Education by signing up for our newsletter.