For instructional designers, the arrival of generative AI has created an uncomfortable question – if technology can generate outlines, draft learning content, suggest assessments, and assemble courses, where does that leave the people who have traditionally done this work?

The more useful question is not whether AI will replace instructional designers, but how AI for instructional design can change where they spend their time and expertise.

Modern L&D teams are being asked to produce more content, respond faster to changing business needs, support increasingly diverse learners, and maintain quality across growing course portfolios. Yet instructional designers still spend significant time on tasks such as structuring first drafts, rewriting SME material, creating content variations, and moving information into course-authoring formats.

AI-assisted authoring can reduce some of that workload. What it cannot replace is the judgment required to decide what learners actually need, how they should learn it, and whether the resulting experience will make a meaningful difference.

Traditional course development often involves a long chain of activities. An SME provides information, an instructional designer interprets it, a course structure is developed, content is drafted, assessments are created, and the material moves through multiple review and development cycles.

This process exists for good reason. Quality learning requires analysis, context, and careful decision-making. The problem arises when skilled instructional designers spend too much of their capacity on repetitive production work rather than instructional thinking.

AI changes that balance.

An instructional designer can use AI to accelerate early-stage course planning, generate a starting structure, explore alternative ways to explain a concept, or draft assessment questions for review. Instead of beginning every task with a blank page, designers can begin with something they can critique, refine, reorganize, or reject.

That distinction matters. AI does not need to make the instructional decision to be valuable. It can help the instructional designer reach a better decision more efficiently.

Instructional design has never been about writing slides or converting documents into courses. At its best, it involves understanding performance problems, identifying appropriate learning objectives, selecting meaningful practice opportunities, anticipating learner misconceptions, and connecting learning to real-world application.

Those responsibilities become more important when AI enters the workflow.

AI can generate great content that is relevant and instructionally useful. However, it may not understand why one example will resonate with a particular workforce while another will confuse learners. It cannot independently determine whether a training request is actually a learning problem and whether a course is even the right intervention.

This is where instructional designers move further into the role of learning architects. They provide context, challenge assumptions, evaluate AI-generated material, and shape raw information into experiences designed around learner and organizational needs.

For teams exploring this approach, the AI-assisted course creation capabilities within Abara Author show how AI, built with instructional design as the core, can support different stages of authoring while leaving room for authors to review and customize what is created.

Consider a hypothetical scenario. A training company needs to develop a new manager onboarding course based on learning goals and SME input.

Rather than manually drafting every section from scratch, an instructional designer could use AI-assisted authoring to create an initial course structure and content draft. The designer can then examine whether the sequence makes sense, rewrite explanations, introduce realistic workplace scenarios, adjust the level of complexity, and ensure assessments test meaningful application rather than simple recall.

The first draft becomes faster to produce, but the instructional designer still owns the learning experience.

This model also changes how teams think about automation. Not every project requires the same degree of AI involvement. Some courses need close manual control, while others may have clear objectives and established source material that make greater automation practical.

Abara Author, for example, supports manual, semi-automated, and fully automated authoring workflows. This gives teams options for deciding where AI assistance makes sense rather than forcing every project through the same development process.

For instructional designers, flexibility may be more valuable than maximum automation.

As AI takes on more production-oriented tasks, several instructional design capabilities become more important, not less.

Critical evaluation is one of them. Designers need to recognize when generated content sounds convincing but lacks appropriate context. They must also decide what to remove, not just what to add.

Learning strategy is another. AI may help generate content around a stated objective, but defining the right objective requires understanding the audience, desired performance, organizational environment, and the reason the training exists.

SME collaboration also remains essential. AI can help transform expert knowledge into drafts, but SMEs provide organizational context, technical accuracy, exceptions, and practical realities that generic generation cannot reliably infer. Abara Author itself notes that AI-generated content still requires human subject matter expertise for verification in its guidance on AI-generated course content.

Finally, designers increasingly need AI literacy. The skill is not simply writing better prompts. It is knowing what to delegate to AI, what to review carefully, and what should remain a human-led decision.

Introducing AI into course development should not mean removing review stages. Instead, teams need to reconsider what happens within those stages.

A practical workflow can begin with human-defined learning needs and objectives. AI can then assist with course planning, first drafts, interactivities, scenarios, or assessments. Instructional designers and SMEs review the outputs for accuracy, instructional relevance, tone, and context before the course moves toward final development and publishing.

Governance also matters. Organizations should establish clear expectations around source material, factual verification, sensitive information, brand standards, accessibility, review ownership, and approval.

This creates a healthier relationship with AI. Instead of asking, “Can AI build this course for us?” teams can ask, “Which parts of this course should AI accelerate, and where does human expertise create the most value?”

For L&D leaders, the opportunity is not simply to produce more courses with fewer people. Treating AI purely as a cost-cutting mechanism risks overlooking its more strategic value.

The bigger opportunity is to redirect specialist capacity.

If instructional designers spend less time assembling routine first drafts, they can spend more time analyzing learning needs, working with SMEs, improving practice activities, strengthening assessments, reviewing learning quality, and addressing more complex business problems.

Training companies may similarly use AI-assisted workflows to handle repetitive production more efficiently while preserving expert attention for projects where instructional depth and customization matter most.

The question for leaders, therefore, is not how much work can be automated. It is how automation can create more room for the work that should not be automated.

The rise of AI for instructional design does not make instructional designers less relevant. It raises the value of the capabilities that have always separated thoughtful learning design from simple content production.

For L&D teams, the goal should not be to choose between human expertise and AI. It should be to design a workflow in which each does what it does best.