Will AI Replace Instructional Designers? My Honest Answer

No. AI is not replacing instructional designers. It is replacing instructional design tasks, and it disproportionately rewards the designers who learn to direct it well. According to Devlin Peck's ID Hiring Manager Report, 92.1% of hiring managers said AI would impact their learning team within 12 months, yet 89.2% said AI is unlikely to reduce the size of that team. Employers expect AI to change the work. They are still hiring humans for the judgment.

That answer needs to be earned, though, every other article on this topic is reassurance without specifics. A real set of tasks is disappearing from the instructional designer's plate right now, and some roles are more exposed than others. I get this question constantly. Whenever we review a new AI capability in the weekly workshops I lead, someone asks, "will instructional designers still be needed if AI can do this?" My answer is always the same: this makes you more capable, and it makes your judgment more valuable. The rest of this article shows you the data behind that answer, names the tasks that are actually going away, and tells you what to build instead. If you want the practical companion to this career question, see how instructional designers are using AI.

What does the data actually say about AI and instructional design jobs?

The labor data points toward transformation with a pay premium for AI skills, not contraction. Here are the numbers that matter from three different sources.

Statistic	Figure	Source
Hiring managers who said AI would impact their learning team within 12 months	92.1%	Devlin Peck's ID Hiring Manager Report
Hiring managers who said AI is unlikely to reduce the size of their learning team	89.2%	Devlin Peck's ID Hiring Manager Report
Hiring managers' teams already using AI in instructional design tasks	48.5% (another 35.6% planned to within a year)	Devlin Peck's ID Hiring Manager Report
Average wage premium for workers with AI skills, across all roles	62%, up from 57% the prior year	PwC 2026 Global AI Jobs Barometer
Growth in job postings requiring AI skills vs. the total job market	69% vs. 9%	PwC 2026 Global AI Jobs Barometer
Projected employment growth for training and development specialists, 2024 to 2034	11%, much faster than average	U.S. Bureau of Labor Statistics
Projected annual openings for training and development specialists	About 43,900	U.S. Bureau of Labor Statistics

Read those numbers together and the picture is consistent: employers expect AI to reshape the work, they do not expect it to shrink their teams, and the market pays more for people who can use it.

Which instructional design tasks is AI already doing?

As of July 2026, AI handles most first-draft production work: topic research, outlines, storyboard drafts, scripts, quiz questions, alt text, and image generation. This is the part vendor articles tend to soften, so let's be specific. i4cp's task-by-task analysis of the instructional designer role, written by senior research analyst Tom Stone, mapped generative AI against the common tasks of the job back in 2023, and the capabilities have only expanded since.

Instructional designers who use AI will replace instructional designers who do not use AI.Tom Stone, i4cp

Here is the current task-by-task picture. Capabilities reflect mainstream AI tools as of July 2026.

ID task	What AI does today	What the human still owns	Exposure
Topic research	Summarizes sources, surfaces key concepts, compiles findings in minutes	Verifying accuracy, choosing what matters for this audience	High
Course outlines	Generates solid draft outlines from a brief	Sequencing for the real performance gap, cutting what does not serve it	High
Storyboards and scripts	Produces usable first drafts of screens, narration, and dialogue	Voice, nuance, and whether the scenario reflects how the job actually works	High
Quiz and assessment generation	Drafts question banks in any format from source content	Deciding what evidence of competence actually looks like	High
Accessibility (alt text, captions)	Drafts alt text and captions at scale	Accuracy review, meaningful descriptions for complex visuals	High
Image and media generation	Strong for scene-setting and mood imagery	Depicting specific processes and procedures accurately	Medium
SME interviews	Drafts question lists, summarizes transcripts	The conversation itself: trust, follow-ups, reading what is not said	Medium
Needs analysis	Organizes data, drafts analysis documents	Diagnosing whether training is even the answer	Low
Instructional strategy	Suggests options from a brief	Selecting and defending the right approach for the constraint set	Low
Stakeholder management	Drafts emails and summaries	Negotiation, pushback, alignment, and trust	Low
Measurement and evaluation	Crunches data, drafts reports	Defining success measures and owning the business result	Low

The interactive checker needs JavaScript. The full picture is below: every task bucket with its exposure rating, and the four skills the checker ranks.

How AI exposure breaks down across common ID tasks
Task	Exposure	Why
Content drafting and storyboarding	AI is absorbing this	First drafts, outlines, and storyboards are what current models do best. The paid work is moving from writing the draft to directing and judging it.
Slide and module production	AI is absorbing this	Rapid-authoring modules and slide builds are template work, and template work is what gets automated first.
Visuals, media, and graphics	Pressure is building	Generation is fast and cheap now, but accuracy, licensing, and brand fit still need a trained human eye.
SME and stakeholder work	Durable for now	Getting knowledge out of busy experts and keeping stakeholders aligned is trust work. Models do not build trust.
Analysis and measurement	Pressure is building	AI will crunch the numbers for you. Deciding what to measure and defending it to leadership stays human.
Strategy and consulting	Durable for now	Framing the business problem and choosing what not to build is judgment work. That is where the durable value sits.

The four skills the checker ranks

AI direction and orchestration. Stop competing with the tools and start directing them. The IDs who scope, prompt, and assemble AI output ship faster than whole teams that draft by hand.
Evaluating AI output. Someone has to catch the confident nonsense before learners see it. Instructional soundness, accuracy, and accessibility review of AI output is already a named responsibility on teams that adopted early.
Measurement and evaluation. When content is cheap, proof gets expensive. IDs who can show whether training changed behavior become harder to cut, not easier.
Simulation and scenario design. Realistic practice is the part of the craft AI raises the ceiling on instead of replacing. Deciding what learners should practice, and making it feel real, stays a human call.

Where does most of your week actually go? Pick every bucket that takes real hours.

Which role is closest to yours?Where is your org on AI?

Pick at least one task bucket first.

Which parts of the job can AI not replace?

The diagnostic and judgment layer: figuring out whether training is even the answer, negotiating with stakeholders, and owning business outcomes. Notice that every low-exposure row in the table above is a decision, not a deliverable.

There is experimental evidence for this. Dr. Philippa Hardman's replacement experiment had around 200 instructional designers blind-score work produced three ways: an experienced ID working alone, a novice using AI, and an experienced ID using AI. On tasks like writing learning objectives, the expert-plus-AI combination came out on top. Expertise did not become irrelevant when AI entered the picture. It compounded.

My take on why this matters for your job security: if you are out here creating real business impact, it is going to be much harder to justify replacing you. If you are just creating pretty eLearning, that becomes an easy choice once AI can match the output. The designers who tie their work to performance outcomes, and who can show the ROI of practice-based training and other measurable results, are the ones whose judgment gets more valuable as production gets cheaper.

How is the instructional designer role changing?

The role is shifting from content producer to what I would call a human expert orchestrator: someone who directs AI agents across many tasks at once, with judgment, creativity, and operating speed as the differentiators.

I can show you what that looks like, because it is how I work now. As of mid-2026, my eight-hour workday is mostly spent directing multiple Claude Code instances, each working on a different project and each able to deploy subagents of its own. On a given day, one might be helping me complete an enterprise security review, another building a document package, another generating a learning experience for a client, and another working through lead activity and sales development. That is one business. I often have 5 to 10 agents running at once, managed through cmux, a multiplexer that lets me run and monitor them in parallel. The work spans everything from document production to designing and developing eLearning to building automations.

Here is the part that answers the replacement question: If I told AI to judge its own output and just "make it better, don't stop until done," we would end up with horrific AI slop. It is almost cliche at this point, but the more you use AI, the more you realize how valuable your judgment and direction actually are. The fatal mistake in this workflow would be publishing and sending everything without reviewing it.

Full disclosure on why I live in this workflow daily: devlin.ai is my company. It is an AI text and voice conversation simulation builder I launched in 2026. You describe a scenario in plain English and get a working simulation that embeds in Storyline, Rise, or any LMS. The instructional designers building in it often have an initial reaction of "wow, this is too easy, why am I needed?" Then, as they try to build the ideal simulation for their use case, they realize another person would not be able to do it anywhere near as well (or that their design skills are still more necessary than they initially thought). Their judgment is what separates a great result from a mediocre one. That mirrors Hardman's finding exactly, just in practice instead of an experiment.

Which instructional designers are most at risk?

Production-only, template-driven roles, where the entire job is converting SME slide decks into modules, carry the highest exposure. A single "you're safe" verdict would be dishonest to the field, so here is the segmentation:

Order-taker production IDs (high exposure). If the job is "take this deck, make it a module," AI already does most of it. These roles will consolidate.
Full-cycle instructional designers (transformed, not replaced). Designers who run analysis through evaluation keep the job, but the production middle of their week gets automated and the judgment layers expand to fill it.
Performance consultants and simulation designers (growing demand). The people who diagnose problems, design practice experiences, and measure outcomes get more valuable, because AI makes their solutions cheaper to build and their judgment scarcer by comparison.

My view is that the gap between AI power users and AI-avoiders is widening fast, and it compounds. There is still a real distance between what an individual practitioner can learn to do and what most organizations even realize is possible. If you can be the instructional designer who introduces your team to an agent-driven workflow, with subagents handling a variety of common ID tasks, the opportunity is enormous.

Aspiring IDs should note one more finding. The PwC 2026 Global AI Jobs Barometer found that AI-exposed entry-level roles are now seven times more likely to require traditionally senior-level skills such as judgment and leadership. Those roles grew 35% since 2019 while other entry-level roles declined. Entry-level does not mean judgment-free anymore. That raises the bar for breaking in, and it rewards people who show up with judgment skills already visible in their portfolio.

What skills should instructional designers build now?

Build the skills that compound with AI rather than compete with it. In priority order:

Directing and evaluating AI output. Not casual prompting: agentic workflows, delegation to subagents, and rigorous review of what comes back. This is the orchestrator skill set from earlier, and it is the widest gap between practitioners and organizations right now.
Scenario and simulation design. AI opens up types of learning experiences that were not practical before, like AI role-play simulations for conversation practice that add a real-time practice layer to existing training. Someone has to design those experiences well, and AI cannot judge its own realism.
Measurement and evaluation. When production gets cheap, proving impact becomes the scarce skill. Designers who can define success measures and demonstrate results own the conversation with the business.
Business consulting. Diagnosing performance problems, pushing back on training requests that will not work, and negotiating with stakeholders. Every one of these is a low-exposure row in the task table, and AI makes people who do them faster, not obsolete.

Each of these gets more valuable as AI improves, because each one is a form of judgment applied to increasingly cheap production.

Is instructional design still a good career?

Yes. The Bureau of Labor Statistics projects 11% employment growth for training and development specialists from 2024 to 2034, much faster than the average for all occupations, with about 43,900 openings per year. If you are weighing the field financially, start with the instructional designer salary data.

There is also a genuine advantage available to newcomers right now: people entering the field can learn AI-native workflows from day one, without unlearning a decade of production habits. Given PwC's finding that entry-level roles increasingly demand senior-level judgment, the transition path that works is the one that builds diagnostic skills, portfolio evidence, and AI fluency together rather than treating AI as an add-on.

One aside for career changers who want structure: Peck Academy is my licensed career school for people transitioning into instructional design. The training covers foundational ID theory, Storyline 360, and AI skills, and an advisory committee of 10 independent industry experts reviews the curriculum to keep it current. The AI skills piece is in there precisely because of everything in this article.

Frequently asked questions

How fast is AI changing instructional design jobs?

Fast on the task level, slow on the role level. Per PwC's 2026 Global AI Jobs Barometer, postings requiring AI skills are growing 69% versus 9% for the overall market, and 92.1% of hiring managers in Devlin Peck's ID Hiring Manager Report said AI would impact their learning team within 12 months. Yet 89.2% of those same hiring managers said AI is unlikely to shrink their team, and BLS still projects 11% growth for the occupation through 2034.

Can AI create an entire course without an instructional designer?

It can generate one. Whether that course changes anyone's behavior is a different question. I predicted early on that full AI-generated courses replacing designers was much further off than the flashy image tools suggested, and that has held up: AI produces plausible drafts, but it cannot diagnose the real performance problem, verify accuracy for your context, or judge its own quality without direction.

Do instructional designers who use AI earn more?

There is no verified ID-specific figure, and you should distrust any site that claims one. The best available data is PwC's 2026 finding of a 62% average wage premium for workers with AI skills, which is a cross-role average across 27 countries, not an instructional design number.

Should I still become an instructional designer in 2026?

Yes, if you build for the field as it is now. BLS projects 11% growth through 2034 with about 43,900 annual openings. The bar has moved, though: PwC found AI-exposed entry-level roles are seven times more likely to require senior-level judgment skills, so learn AI-native workflows and diagnostic skills from the start instead of treating AI as optional.

Will AI replace eLearning developers too?

The same logic applies with higher task exposure. Pure build work, such as assembling screens from a finished storyboard, is being absorbed quickly by AI-assisted authoring. Developers who direct AI, integrate new experience types like practice simulations, and own quality judgment stay valuable; developers whose entire value is manual assembly are the most exposed group in the field.