AI in Instructional Design: The Complete 2026 Guide

AI is now a standard part of instructional design work. The differentiator in 2026 is no longer whether you use it. It is how well AI is integrated into your actual workflow and where you keep human judgment in the loop.

This guide covers what AI can actually do across the design process, how to integrate it without shipping slop, the new skills the job market pays for, and the one capability (AI practice simulations) that changes what instructional designers can deliver.

How are instructional designers using AI in 2026?

The large majority of L&D professionals now use AI, and they use it mostly for speed in design and development work. According to the Synthesia AI in Learning & Development Report 2026, which surveyed 421 L&D professionals in late 2025, 87% of respondents already use AI and only 2% have no plans to adopt it.

Here is where the usage actually concentrates:

Finding	Figure	Source
L&D professionals already using AI	87%	Synthesia AI in L&D Report 2026
Cite speed as the biggest incentive	84%	Synthesia AI in L&D Report 2026
Use AI for voice generation	63%	Synthesia AI in L&D Report 2026
Use AI for content and quiz drafting	60%	Synthesia AI in L&D Report 2026
Use AI for video creation	52%	Synthesia AI in L&D Report 2026
Use AI for translation and localization	38%	Synthesia AI in L&D Report 2026
Planning to adopt AI assessments and simulations	36%	Synthesia AI in L&D Report 2026

Two notes on the data: First, it comes from a vendor (Synthesia), and the report itself acknowledges its respondents skew toward earlier adopters. Second, the direction of travel still matches everything I see in the field.

It also matches how far things have moved. When I surveyed instructional designers back in 2024, only about 24% used AI daily, roughly 29% used it weekly, and 28% never used it at all. The sharpest finding from that survey was the permission gap: 37% of instructional designers did not know whether they were even allowed to use AI in their work. That problem has not gone away. In Synthesia's 2026 data, security concerns (58%) and legal constraints (41%) are now among the top blockers to adoption. The pattern is consistent: the tools moved faster than the governance.

What can AI actually do in the instructional design workflow?

AI helps at every stage of the design process, but its value concentrates where the work is repetitive, and its risk concentrates where judgment matters. Use the matcher below to rank these use cases against your own situation, then dig into the ones at the top of your list.

Analysis and research

AI is genuinely good at the front end of a project: synthesizing SME interview transcripts, summarizing source documents, spotting patterns in performance data, and turning a messy pile of inputs into a clear picture of the performance problem. Feed it your raw notes and ask it to surface themes, contradictions, and open questions before your next stakeholder meeting. This is also where AI helps you prioritize: given learner data and business goals, it can help you argue for which learning experiences matter most and which requests to push back on.

The judgment call stays yours. AI can summarize what your SMEs said, but it cannot tell you which SME is wrong.

Design and storyboarding

AI can draft outlines, storyboards, and objective-aligned content sequences in minutes instead of days. The catch is that a first-draft storyboard from a general-purpose model is a starting point, not a complete design. It favors generic structures unless you feed it real constraints: your audience, the performance gap, the format, and the time budget.

I break down where AI fits in the design phase in a dedicated article. And if you are designing scenarios, the design thinking still comes first: scenario-based learning is about finding decisions worth practicing, which no model will do for you.

Content and media development

This is where most teams start, and where the tooling has matured fastest. AI drafts scripts, generates quiz questions from objectives, produces narration without a recording booth, and creates course imagery from a prompt. Authoring tools now ship this natively: Articulate's AI Assistant is live inside Storyline 360 and Rise 360, so drafting, summarizing, and image generation happen inside the authoring environment.

An observation from my own work: for a long time, getting reliable, on-brand imagery out of AI required specialized skills, and consistent characters across scenes were basically impossible. That has changed. Modern image tools handle character consistency well now, and it gets better every quarter. If you wrote off AI imagery in 2023, it deserves another look.

Personalized and adaptive learning

AI can adjust content difficulty, recommend resources based on learner behavior, and build individualized paths through material. But we need to be real about where this sits on the hype curve: in the Synthesia 2026 report, 72% of L&D professionals expect personalization to be AI's biggest future impact, while only 24% say they see that value today. Personalization at scale is where the field is heading, not where most teams are. If your LMS promises adaptive learning, ask to see it working before you build your strategy around it.

Assessment and feedback

AI can generate assessment items from objectives, and more usefully, it can evaluate open-ended responses against a rubric and give learners instant, specific feedback. This is what instructional designers have been asking for for decades. Multiple-choice questions were never the bottleneck. The expensive part is evaluating what a learner actually says or writes in their own words, and that is exactly what modern AI evaluation does well. It is also the bridge to the biggest shift in the field, AI simulations (which gets their own section below).

Accessibility and localization

AI makes accessible content dramatically faster to produce: auto-generated captions, audio transcription, alt text for images, readability analysis, and translation into dozens of languages at a fraction of the old cost. Translation is already mainstream, with 38% of L&D teams using AI for it per the Synthesia 2026 report. Treat AI output here as a strong first pass. Captions and translations still need human review, especially for technical terminology and anything compliance-related.

How do you integrate AI into your workflow instead of just collecting tools?

Pick one recurring task, define exactly what AI drafts and what you review, and make the review step non-negotiable. That is the whole playbook. The field's conversation has moved past "which tools should I try?" and onto how you actually work alongside AI day to day. Dr. Philippa Hardman's State of Instructional Design research has been tracking this shift in the ID role for years, and the pattern is clear: the designers getting outsized value are not the ones with the longest tool list. They are the ones who redesigned a workflow around AI and kept quality control.

Here is how I actually work, as of July 2026. I build devlin.ai and run my content automation using Claude Code, Anthropic's terminal-based coding agent, with the most capable models available. That used to mean the Claude Opus family; right now it means Fable 5, and it will mean something else next year, which is exactly the point. The workflow is stable even though the models churn: AI does the heavy drafting and execution, and I review everything before it ships. Nothing goes out one-shot.

If you want to build the same kind of setup for your ID work, here is the sequence I recommend:

1. Pick one recurring task

   Choose something you do weekly that eats hours: storyboard drafts, quiz generation, SME interview synthesis, alt text. One task, not five. You want a tight feedback loop.

2. Clarify your org's AI policy first

   Before you touch company content, find out what is approved, what is restricted, and what is under review. If no policy exists, ask your manager in writing. This step is boring and it is the one that protects your job.

3. Define the handoff

   Write down what AI produces and what you review. For a storyboard: AI drafts structure and placeholder content; you own the scenario logic, the tone, and every factual claim.

4. Make review non-negotiable

   Every AI output gets human eyes before anyone else sees it. Build the review time into your estimate. AI plus review is still dramatically faster than doing it all by hand.

5. Measure, then expand

   Track the time saved on that one task for a month. Then, and only then, add the next task. A documented win also gives you leverage when you ask for tool budget.

If your integration question is bigger than your own workflow (like getting a whole team or organization on board), that is a change-management problem, and I cover it in getting an AI simulation rollout approved and adopted.

What are AI practice simulations, and why are they the biggest shift?

AI simulations let learners practice real conversations in their own words, by voice, and get evaluated automatically against criteria you define. In my view this is the first genuinely new learning experience category that AI has enabled. Everything else on this page makes existing work faster. This makes new work possible.

Here is the before and after from my own career. Building branching scenarios by hand in Storyline used to take me weeks, and the ambitious ones took months: mapping every branch, writing every response, wiring every trigger. Today I can integrate AI-powered practice into a Storyline course in a few hours, and the result is more engaging than the branching version because learners respond in their own words instead of picking from three canned options.

The bigger reason that this is important: realistic conversation practice used to require human role-play facilitators. For most teams, that was simply out of budget, so the practice never happened and people rehearsed difficult conversations for the first time on real customers and real employees. AI removed that constraint. And voice is what makes it work. Speaking through a tense conversation out loud is a completely different rehearsal than typing (or, worse, selecting multiple-choice options), because the hesitation, the tone, and the recovery are the skill.

Full disclosure: this is the problem my company works on. devlin.ai is an AI text and voice simulation builder I launched in 2026, and if you are weighing it against other role-play options, I wrote a comparison of Synthesia Roleplay Sessions vs devlin.ai. You describe a scenario in plain English and get a working simulation that embeds in Storyline, Rise, or any LMS, with AI evaluation, transcripts, and scores reporting back to the course. It supports voice and text in 16 languages, and there is a Storyline 360 variable bridge, so a simulation can update Storyline variables like score, pass/fail, or even a character's mood to drive states. More than 1,500 instructional designers are building with it, and the two things they tell me they value most are the time savings and the fact that this category of experience simply was not available to them before. I obviously have a stake here, so weigh my enthusiasm accordingly, but I built the company because I saw the field moving towards this new format.

If you want to go deeper: here is how AI training simulations work end to end, how AI role-play simulations handle soft-skills practice, and if you need to justify the budget, the ROI case for practice-based training.

What AI skills do instructional designers need in 2026?

AI competency has become a hiring filter, and the market pays a measurable premium for it. According to PwC's 2026 Global AI Jobs Barometer, which analyzed over one billion job ads across 27 countries, the average wage premium for workers with AI skills reached 62%, up from 57% the year before, and jobs requiring AI skills are growing roughly eight times faster than the overall jobs market. Just as telling: the same research finds AI-exposed roles increasingly demand judgment, creativity, and leadership. The premium is not for button-pushing. It is for direction.

For instructional designers specifically, the skills that matter are:

• Prompt fluency. Getting useful output on the first or second try because you supply context, constraints, and examples.
• Output evaluation. Catching hallucinations, bias, and generic filler before learners do. This is instructional design judgment applied to a new input source.
• AI-era assessment design. Designing rubrics and evaluation criteria that an AI can score against reliably, especially for open-ended and conversational responses.
• Data interpretation. Reading the analytics that AI-enabled tools produce and turning them into design decisions.
• Knowing what to keep human. The meta-skill: deciding which parts of the work should never be delegated, like stakeholder relationships and the final quality call.

These skills also move salary conversations, and it helps to know your baseline. Here is what instructional designers actually earn and how to negotiate from data.

If you are transitioning into the field and want structure while you build these skills, that is what Peck Academy, my licensed career school, is for. The training covers foundational ID theory, Storyline 360, and AI skills, and students build an AI-integrated flagship project in Storyline 360 with extensive feedback, plus a portfolio website. I own it, so factor that in, but AI skills are in the curriculum precisely because of the hiring data above.

What are the risks of using AI in instructional design?

The biggest risk in 2026 is not robots taking jobs. It is AI slop: unreviewed, generic content shipped at scale (with your name on it).

I use AI heavily every day, so let me be specific about where it fails, from my own attempts:

• One-shotting does not work. Every time I have tried to ship an AI output with no human review, quality problems surfaced. Sometimes subtle, sometimes embarrassing. The review pass is not optional.
• One-click course generators produce slop. Nearly every learning vendor now ships a "generate a course from this document" button. I have tried them. The output is structurally fine and instructionally hollow: no real scenario, no real practice, no point of view. It is an infodump, not a learning experience.
• On-brand imagery was hard until recently. I could not get consistent characters or brand-faithful visuals without specialized skills. Newer image models have largely fixed this, but it is a useful reminder that "AI can do X" claims come with a "how reliably?" question attached.

Beyond my own list, the systemic risks are real and worth guardrails:

• Hallucination. Models state false things confidently. Any factual claim in learner-facing content needs verification against a source you trust.
• Bias. AI reproduces the biases in its training data, in scenarios, imagery, and example names. Review with that lens explicitly.
• Data privacy. Never paste learner PII or confidential company material into tools your organization has not approved. You are in good company being careful here: per the Synthesia 2026 report, 59% of L&D teams avoid using personal or sensitive data with AI entirely, and security is the top adoption blocker at 58%.
• The permission gap. If you do not know your org's AI policy, that is your first risk to close. Ask before you paste, and if you do client work, check whether your contracts require you to disclose AI use.

One more risk category deserves its own scrutiny: tools that claim to evaluate learners. If an AI is scoring people, its reliability is a fairness question, not just a quality question. Here is how to evaluate whether an AI simulation tool's scoring is reliable before you trust it with real learners.

What are the best AI tools for instructional design?

The best tool depends on the job: a general-purpose assistant covers most day-to-day tasks, and specialized tools earn their place for voice, video, imagery, and practice simulations. For the current breakdown of what each tool is actually good at, where it falls down, and how to build a stack, see my guide to the best AI tools for instructional design. And check the date on any tool roundup you read, including that one, because AI tool lists rot faster than any other format.

Will AI replace instructional designers?

No. That is my honest answer to whether AI will replace instructional designers, and the full case, with the labor data and my survey findings behind it, lives in that guide. The short version: AI is replacing tasks, not designers. It removes production work, not the judgment about which conversations are worth practicing and what good performance looks like, and it disproportionately rewards the designers who learn to direct it well.

Where is AI in instructional design headed?

Toward agentic, integrated systems: AI that acts across your tool stack rather than living in a chat window you copy-paste from. The early signals are already in the data. Per the Synthesia 2026 report, only 47% of L&D professionals believe the LMS will still be the backbone of their learning ecosystem in three years. Agentic AI exploration is concentrating on AI tutors (49%) and coaching and mentoring (43%), and assessments and simulations lead planned adoption at 36%, the fastest-growing pilot category.

Here is my grounded read on what to do about it:

• Do this quarter: get your one-task workflow integration running, close your permission gap, and build or pilot one AI practice experience so you understand the category firsthand.
• Watch, but do not chase: agentic multi-tool workflows, AI tutors embedded in the flow of work, and the post-LMS ecosystem debate. These are real trends, and they will still be maturing next year. You do not need to re-platform to stay current. You need working AI habits, because every one of these future systems assumes a designer who can direct AI and judge its output.

Frequently asked questions

Will AI replace instructional designers?

No. AI is replacing tasks (drafting, production, first-pass content) rather than designers. According to PwC's 2026 Global AI Jobs Barometer, roles where AI amplifies expert judgment are growing faster and paying more than roles AI automates. The designers at risk are the ones who neither integrate AI nor develop the judgment skills it cannot replicate.

Do I need coding skills to use AI as an instructional designer?

No. Nearly every AI use case in this guide (drafting, storyboarding, media generation, quiz creation, simulation building) works through plain-English prompts. Coding helps if you want to automate workflows across tools, but it is an amplifier, not a prerequisite.

Is it OK to use AI at work if my organization's policy is unclear?

Ask first, in writing. When I surveyed instructional designers in 2024, 37% did not know whether they were allowed to use AI at work, and security remains the top organizational blocker in 2026 data. Until you have clarity, never put company content, learner data, or anything confidential into an unapproved tool. Practicing on your own non-confidential projects is a safe way to build skills in the meantime.

What is the fastest way to start using AI in instructional design?

Pick one recurring task that eats hours each week, such as storyboard drafts or SME interview synthesis. Define what AI drafts and what you review, run it for a month, and measure the time saved. One integrated task beats ten tools you opened once.

Will AI-built courses pass quality review?

Not without human work. One-click AI course generators produce content-shaped output with no real scenario design or point of view, and reviewers increasingly recognize it. AI-assisted courses pass quality review when a designer owned the design decisions, verified the facts, and reviewed every output. AI-generated courses that skipped those steps usually do not deserve to pass.

Next steps

Your next step depends on where you are. If you are a practicing instructional designer, run the workflow integration steps above on one task this week; that single habit compounds more than anything else on this page. If practice and rehearsal are the gap in your programs, start with how AI training simulations work and build a small pilot. And if you are still transitioning into the field, focus on the skills section: the market is telling you, with a measurable wage premium, that AI competency belongs in your portfolio from day one.

The tools will keep churning. The designers who thrive are the ones who build the judgment to direct them.