As Instructional Designers, AI tools like ChatGPT are becoming more integrated into our workflows. While this is happening, many of us are asking: How can we use these tools and what are their limitations? With AI increasingly assisting in tasks like content development, learning objective generation, and feedback summarization, we need to understand how these tools align with the levels of cognitive complexity we aim to foster in learners—and where human instructional designers remain irreplaceable.

This article uses one of Instructional Design’s fundamental and reliable tools as a framework to critically evaluate the role of Large Language Models (LLMs) in educational technology: Bloom’s Taxonomy. The purpose of using Bloom’s is twofold: for one, it is easier to explore a new technology using a framework we are familiar with; additionally, Bloom’s Taxonomy gives us the ability to meet LLMs where they are, which is a great way to integrate a coauthor or collaborator, even an artificial one.

Applying Bloom’s to LLMs: A Practical Breakdown

1. Remembering

(Recall of facts and basic concepts)

LLMs excel at pulling up definitions, frameworks, historical facts, and terminology—often faster than a Google search. Need a quick explanation of SCORM vs. xAPI? No problem.

Strength: Instant access to surface-level knowledge.

Limitation: Factual hallucinations are possible, especially with niche or emerging topics.

Design Use Case: Great for generating knowledge checks, glossaries, or quick refreshers—but double-check accuracy.

Summary: This level is well supported by LLMs and requires minimal oversight.

2. Understanding

(Explain ideas or concepts)

ChatGPT can explain instructional strategies, translate complex SME jargon into plain language, and paraphrase source content into new tones or modalities.

Strength: Helps rephrase and contextualize concepts.

Limitation: Doesn’t “understand” the way humans do—lacks intention or judgment.

Design Use Case: Useful for drafting overviews, summaries, or learning objectives, with your editorial review to ensure clarity and alignment.

Summary: This level is also well supported, but moderate oversight ensures instructional alignment.

3. Applying

(Use information in new but familiar situations)

Need to apply Universal Design for Learning (UDL) or adult learning theory to a draft lesson? LLMs can offer a solid starting point.

Strength: Good mimicry of instructional structures and use cases.

Limitation: No real-world awareness; lacks learner empathy or contextual judgment.

Design Use Case: Use it to jumpstart prototypes, then refine through learner analysis and empathy mapping.

Summary: Supported with limitations—LLMs can assist, but human designers must guide contextual application.

4. Analyzing

(Draw connections, compare, break down ideas)

LLMs can generate basic comparisons, contrasts, and SWOT-style breakdowns. They can also dissect lesson components at a structural level.

Strength: Rapid identification of patterns and relationships.

Limitation: Analysis tends to be surface-level or generic—without context, it misses subtleties.

Design Use Case: Helpful for initial outlines or decision trees, but needs human refinement for real instructional rigor.

Summary: Partially supported—may assist with structure, but deeper analytical work requires human input.

5. Evaluating

(Critique, justify decisions, make judgments)

ChatGPT can simulate evaluation (e.g., provide pros and cons or suggest improvements), but it's pattern-based, not value-driven.

Strength: Language that looks evaluative, useful for prompting reflection.

Limitation: No actual criteria-based reasoning or expertise.

Design Use Case: Use cautiously for early feedback templates, but rely on human judgment for assessments and recommendations.

Summary: Not reliably supported—outputs may appear evaluative but lack substance.

6. Creating

(Generate new content, ideas, or solutions)

This is where LLMs appear to “wow” us. They can draft entire lesson plans, scenarios, reflection prompts, or analogies on demand.

Strength: Fast ideation and content scaffolding.

Limitation: Outputs are remix, not innovation—often lack originality, audience alignment, or instructional precision.

Design Use Case: Excellent for brainstorming and first drafts, but human creativity and context bring the magic.

Summary: Supported with limitations—great for ideation, but not for finished instructional products.

Summary: LLMs Through the Bloom’s Lens

What Should Instructional Designers Do with This?

Understanding the boundaries and capabilities of LLMs helps us:

Design Responsibly: Use AI as a co-pilot, not an autopilot.

Communicate Clearly: Help clients and stakeholders see where your expertise is essential.

Collaborate Efficiently: Delegate low-stakes or first-pass tasks to LLMs, freeing you to focus on deeper cognitive design.

Final Thought: Let Bloom’s Guide More Than Learning

LLMs are powerful tools in the hands of a thoughtful Instructional Designer—but they’re still just that: tools. Let Bloom’s Taxonomy guide not just your learners’ growth, but your approach to partnering with AI. Embrace the possibilities, stay alert to the limits, and always keep the human at the center.

Image credit: Created using ChatGPT

Want to Learn More? Connect with your institute's D2L Customer Success Manager or Client Sales Executive, or reach out to the D2L Sales Team for more information about how Learning Services can support you on your learning journey.