A Practical SAMR + AI Framework for Instructional Design

Educator and media theorist Neil Postman argued that a new technology does not simply add to or subtract from an environment; it changes everything.^Footnote1 Generative artificial intelligence (GenAI) illustrates this transformation, reshaping the design and evaluation of teaching and learning activities. The 2025 EDUCAUSE AI Landscape Study found that only 22 percent of students use GenAI to learn discipline-specific workforce skills, despite it being a priority for most institutions.^Footnote2 This gap between institutional priorities and students' use presents an opportunity for educators and instructional designers. By responding to this technological shift and evolving learning activities, teaching and learning professionals can preserve meaningful learner cognition and authentic engagement.

Traditionally, educators designed activities centered on brainstorming, analysis, and synthesis—tasks that engaged critical and creative thinking and provided evidence that students were constructing and organizing knowledge. However, the ability of AI tools to generate outputs that mirror these processes poses a significant challenge to maintaining academic rigor. How can instructors preserve rigor and authenticity when GenAI tools can generate similar outputs?

To respond to this shift, instructors must change what they ask of students. In this new instructional context, GenAI acts as tutor, assistant, partner, provocateur, or co-creator. With intentional learning design, educators can use GenAI to develop a deeper understanding and authentic engagement.

A practical framework for integrating GenAI into academic coursework was developed to facilitate this work. The framework supports disciplines where learning is demonstrated through student reasoning (e.g., natural sciences, mathematics, humanities, social sciences, and many professional programs). Drawing on the SAMR Model of technology integration and Bloom's Taxonomy, the framework aligns levels of technology use with cognitive demand to support strategic instructional redesign. The approach uses a five-step process anchored by the SAMR + AI Matrix, a learning design tool created to help instructors determine exactly how they want GenAI to impact student work.

The Impact of Technology on Learning

An effective transition into designing in this new ecosystem should be supported by a method for measuring the effect a new technology will have on the learning experience. The SAMR Model, developed by Reuben Puentedura, is a simple and powerful framework that categorizes the impact of technology into four levels:^Footnote3

• Substitution: The technology replaces a traditional tool with no functional improvement.
• Augmentation: The technology replaces and improves function or efficiency.
• Modification: The technology enables a significant redesign of the task.
• Redefinition: The technology enables the creation of new tasks previously inconceivable.

The SAMR Model helps instructors evaluate whether a technology enhances or transforms a learning activity (see Figure 1).

Figure 1. The SAMR Model

GenAI complicates evaluation because it can produce artifacts that are traditionally used as evidence of student thinking and learning along each of Bloom's levels. Considering the SAMR Model and Bloom's Taxonomy together can help instructors decide how GenAI can support learning at each cognitive level. To guide the redesign of learning experiences, the relationship is synthesized in the SAMR + AI Matrix. Presented as an interactive table, the Matrix is a structured learning design tool that brings together cognitive demand (Bloom), technological impact (SAMR), and GenAI functionality.

5 Steps to Enhance or Transform Instruction with SAMR + AI

These five steps can help educators deliberately design for the presence of GenAI in the learning experience.

Step 1: Reflect on Your Current Practices

Before jumping into a course redesign, evaluate where you are. Ask yourself the following questions:

• At which SAMR level does GenAI currently impact my course?
• How, if at all, am I currently supporting my learners in using GenAI responsibly with my content?

This reflection can help determine whether a transformation is warranted or a simple enhancement might suffice.

Step 2: Select One Instructional Component

Select an assignment, activity, or assessment that meets one of the following criteria:

• You use it regularly and feel comfortable deconstructing it.
• It is repeated or central to your course.
• You suspect it could be more effective at helping students process the content.
• Students consistently struggle with it or have misconceptions about it.

To illustrate how the redesign works in practice, consider a common assignment, such as a research paper in which students use scholarly sources to support an argument.

Step 3: Identify the Bloom's Level of Your Component

Each learning task should align with a level on Bloom's Taxonomy.^Footnote4 From basic recall to invention, identify the primary Bloom's level of your chosen instructional component. A learning outcome for a research paper assignment might require students to critically examine competing scholarly arguments and defend a reasoned, evidence-based position. Because students must critique arguments and support their claims using evidence, this outcome aligns with Bloom's Evaluate level.

Step 4: Integrate the AI with Intention

The SAMR + AI Matrix can guide the integration of GenAI into learning tasks. Here's how to use it:

1. Locate the Bloom's level: Find the row that matches your chosen component.
2. Select the level of impact: Move horizontally to the desired SAMR level.
3. Extract the GenAI action: Use the SAMR + AI Matrix to copy the GenAI action in one click.
4. Employ the prompt templates in two rounds (see Prompt Template section below).

In the research paper example, the instructor may decide to completely transform the assignment—reimagining the task and inviting students to produce an artifact other than a traditional paper. In this case, the instructor would begin at the Evaluate row of the SAMR + AI Matrix and move to the Redefine column. The intersecting cell provides a design strategy aligned to that cognitive level and technological impact. That strategy is then inserted into the structured prompt template to generate domain-specific assignment ideas that preserve the intended level of learning.

Prompt Templates ^Footnote5

Round 1

I am designing an activity/assignment for a [your subject] course at the [undergraduate/graduate/doctoral] level.

I want the students to use GenAI to support this activity or assignment in the following way: [paste content from the selected cell in the SAMR + AI Matrix].

The topic students are studying is: [insert your specific topic].

The learning goal is for students to: [insert Bloom's learning objective in your own words; e.g., "analyze how major events led to X" or "evaluate competing solutions to Y"].

Generate a list of domain-specific assignment ideas using this task. For each assignment, identify the following:

1. Assignment description
2. GenAI task
3. Student deliverable

Select your preferred assignment from the list of generated ideas, then use the following prompt:

Round 2

I've selected [paste preferred activity from generated list].

Generate:

1. A short assignment description that uses GenAI in this way
2. A sample GenAI prompt students could use
3. A sample output the GenAI might return
4. Suggestions for how to evaluate student work

Step 5: Ensure Active Cognitive Engagement

Students' use of AI should support, enhance, or challenge their thinking, not replace it. Once you've selected the activity to present to the students, evaluate it for cognitive engagement. Despite it being a powerful tool, GenAI, in its current form, is considered to be "narrow" or "weak" AI by computer scientists.^Footnote6 It lacks the human-level ability to learn and apply knowledge across a range of tasks. Given this limitation, students should expect errors, omissions, and overgeneralizations in AI-generated outputs. Requiring students to assess the accuracy, relevance, and precision of such outputs creates a critical opportunity to engage in higher-order cognitive processes.^Footnote7

Determine whether the activity invites the learner to do one or more of the following things:

• Evaluate GenAI output for accuracy, relevance, and precision.
• Capture their own voice, including their thoughts, questions, concerns, and reasoning.
• Make decisions independently and challenge or revise the output of the GenAI.
• Demonstrate what they've learned or where learning broke down.
• Compare their own ideas with AI-generated content and revise thoughtfully.

These indicators help ensure that GenAI functions as a catalyst for deeper cognitive engagement rather than as a replacement for student thinking.

The Moment to Act

GenAI has altered the relationship between task completion and evidence of learning. In response, the SAMR + AI Matrix, when applied alongside disciplinary expertise, offers a structured approach to positioning GenAI as a tool that supports students' preparation to make meaningful contributions in their field. Rather than overhauling an entire course at once, educators can begin with a single instructional component—an assignment, activity, or assessment—identify its Bloom's level, and use the SAMR + AI Matrix to explore how the assignment could be shifted toward more transformative uses of technology. This targeted approach allows instructors to meaningfully support students' use of GenAI in their professional domains.

Notes

1. Neil Postman, Technopoly: The Surrender of Culture to Technology (Knopf, 1992).Jump back to footnote 1 in the text.↩
2. Jenay Robert and Mark McCormack, 2025 EDUCAUSE AI Landscape Study: Into the Digital AI Divide, (EDUCAUSE, February 2025).Jump back to footnote 2 in the text.↩
3. Ruben R. Puentedura, "Transformation, Technology, and Education," Hippasus, 2006.Jump back to footnote 3 in the text.↩
4. "Taxonomies of Learning," Derek Bok Center for Teaching and Learning, Harvard University, n.d.Jump back to footnote 4 in the text.↩
5. ChatGPT-4o, response to "optimizing an instructional design prompt for LLM use," OpenAI, June 27, 2025.Jump back to footnote 5 in the text.↩
6. "Types of AI: Explore Key Categories and Uses,"Syracuse University School of Information Studies, March 26, 2025.Jump back to footnote 6 in the text.↩
7. Hao-Ping (Hank) Lee et al., "The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers," in Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI '25) (Association for Computing Machinery, 2025), art. 1121, 1–22.Jump back to footnote 7 in the text.↩

---

Alyshia Keys-Harris is a Learning Designer at Harvard University.