The Ethics of Designing Training: Should You Build That Course?

The biggest ethical problem in instructional design has nothing to do with controversial course topics. It's that we build courses without any real analysis backing them up, and we rarely check whether they actually worked. Nobody is going to ask you to design a course on how to rob a bank, and unless there's an eLearning black market I don't know about, you won't get hired to teach embezzlement either. Your personal values might steer you away from certain clients or industries, and that's your call to make. It's also not what this article is about.

This article is about the ethics of the profession itself: whether you should build that course at all, and whether anyone will ever measure the result with a framework like the Kirkpatrick model of training evaluation.

As Cathy Moore puts it in Map It, a book I've reviewed in depth:

Instructional design isn't a profession because we don't have a code of ethics. Creating courses on demand is as unethical as prescribing antibiotics on demand. Yet it's so widespread that it has become the standard in our field. (384)

When I first published this article in 2019, I treated that as the final word. But today, codes of ethics do exist. What we still don't have is enforcement, and the course mill keeps humming. Let's take it from the top.

Why is designing training on demand an ethical problem?

Designing training on demand is an ethical problem because it spends learner time, organizational budget, and your own effort on an intervention that analysis would often show can't fix the underlying performance issue.

I've been making the same observation since 2018: someone comes to you, or your manager, or someone higher up the chain, and says "we need a training on X." And 99 percent of the time, the instructional designer complies and goes straight into build mode. No analysis of the actual problem. No question about whether training can even touch it.

Think about what that costs. Your time designing the training. The learners' time sitting through it. The organization's budget bankrolling the whole project. Nobody is likely to die from an unnecessary eLearning course, but it is exactly that: unnecessary.

Pushing training at an issue without conducting a proper analysis is like playing darts in a dark room.

And failing to evaluate the training afterward is like tallying up your score without ever turning the lights on. When we work this way, the odds of success are so low that we usually end up wasting everyone's time, including our own.

To be clear, I'm not saying training doesn't help. Employees can get real value from corporate training when it's relevant to their jobs, especially when it's practice-based. The problem shows up when:

We design courses without identifying what's causing the performance issue.
We push courses at issues that have better solutions.
We force people to take training that won't help them perform better.

Do instructional designers have a code of ethics?

There is no single, enforced code of ethics that governs everyone who holds the "instructional designer" title. Professional bodies do publish codes, though, and in 2023 the field's largest academic association wrote ethics into the definition of the field itself.

The 2019 version of this article claimed flatly that we have no code of ethics. That's no longer a fair summary, so let's correct the record. Here's what actually exists:

Organization	Ethics document	What it covers	Binding on whom
AECT	Code of Professional Ethics	Commitments to individual learners, to society, and to the profession	AECT members only
ISPI	Code of Ethics, plus 10 Performance Standards	Ethical practice in performance improvement, including standards like "Determine Need or Opportunity" and "Determine Cause" before designing solutions	ISPI members and certificants
ATD	Code of Ethics	"The highest possible standards of personal integrity, professional competence, sound judgment, and discretion" in talent development work	ATD members only

On top of that, AECT's board-approved 2023 definition of educational technology opens with "the ethical study and application" of theory, research, and practices. The field is formalizing ethics at the definitional level.

So why does Cathy Moore's provocation still bite? Because none of this is enforced. Membership in these organizations is optional. There's no license to practice, no board review, and nobody loses the "instructional designer" title for churning out useless courses. Doctors who prescribe recklessly face consequences. We don't. In practice, the field still behaves as if no code exists, which means the ethical burden falls on how you, individually, choose to work.

When is training the right solution?

Training is the right solution only when the performance gap is caused by missing skills or knowledge. If people could do the task if their life depended on it (dramatic, I know), then no course would close the gap.

Edmond Manning, writing for Allen Interactions back in 2017, named the usual non-training suspects well: people underperform because there's no reward for doing it right, because environmental obstacles get in the way, because of management problems, or because they're juggling contradictory priorities. A course fixes none of those. It's worth learning to recognize when the problem isn't a lack of training before you open your authoring tool.

Even when knowledge is part of the gap, a course is rarely the cheapest or fastest fix. Performance support tools like job aids, email sequences, and curated resources are routinely overlooked in favor of courses and workshops that cost ten times as much and ship ten weeks later.

How do you find out?

You conduct a needs assessment. Talk to the people doing the work. Look at the data behind the request. Identify the business goal, the behaviors that would move it, and what's actually stopping people from performing those behaviors today. The answer tells you whether you're designing training, recommending something else, or both.

How do you push back when someone demands a course?

You don't refuse. You ask why, surface the data, name the specific risk of proceeding as ordered, and offer to design a more appropriate solution.

We're performance consultants, not order-takers. When a client, SME, or executive comes to us asking for a course, we don't just say yes. We ask why and get to the root of the performance problem.

Here's what that looked like on a recent project of mine. A subject matter expert wanted to cram a large amount of content into the course. Instead of conceding, I told them the content was genuinely valuable, but that our main focus was changing behavior, and that a different approach could help us get there better. Then I showed them examples of what solutions could look like beyond the eLearning courses they were familiar with. We ended up designing a job aid and scenario-based practice opportunities instead. No confrontation required: just a better option they hadn't seen before.

When showing a better option isn't enough, Manning's essay offers the most useful pushback tactics I've seen in print. Adapted with credit:

Name the risk specifically. Not "this might not work," but "if we skip practice activities, new hires still won't be able to handle the top three call types."
Communicate a threat level. Is this a minor waste or a serious liability? Say which, and say why.
Say the hard sentence. Something like: "I'm not convinced this course will achieve the outcomes you want." It's uncomfortable. but it's good to get on the record.
Offer mitigation. If the course must happen, propose the additions that give it a fighting chance: practice scenarios, a job aid, a manager checklist, etc.
Schedule a reassessment. Get a follow-up date on the calendar to look at the results together. Evaluation is your leverage for next time.

What new ethical issues does AI create for instructional designers?

AI adds four ethical duties this debate didn't have in 2019: checking outputs for bias, verifying accuracy in regulated content, disclosing when AI produced or evaluates learning content, and protecting the learner data AI tools collect.

A 2025 systematic review in Frontiers in Education on the ethics of generative AI in education flags data privacy and algorithmic bias among the most pressing concerns, and those map directly onto our work.

Bias and fairness in AI-generated content

Generative tools reproduce the patterns in their training data, including the biased ones. If AI drafts your scenarios, characters, and feedback, you're accountable for reviewing what it produces. "The model wrote it" is not an ethical defense; you shipped it.

Accuracy and accountability in regulated industries

In healthcare, finance, and safety training, a hallucinated detail isn't a quirk. It's a liability that can hurt someone. The same accountability question applies when AI evaluates learners, not just when it writes content. My company, devlin.ai, builds AI text and voice conversation simulations, and our evaluation engine scores learners against rubrics the designer defines. We regression-test that engine, throwing all sorts of transcripts at it until it gives the same scores on the same transcripts every time, because some clients rely on those scores for learner certifications. Getting it wrong has real consequences. And since AI can still make mistakes, I believe human review of failing transcripts will always belong in an ethical setup. Whatever tool you use, the ethics question is the same: how does the vendor keep AI evaluation consistent, and what's the human backstop?

Disclosing AI authorship to learners

Learners deserve to know when they're interacting with a machine. In our real-time AI simulations, we always disclose that the content is AI-generated, and I'm equally open about how I use AI to run my business. If you'd feel awkward telling learners how a piece of content was made or scored, that awkwardness is valuable information for your decision-making.

Learner data privacy

AI-powered learning tools can capture conversations, scores, and detailed performance data. Before you adopt one, know what it collects, where that data lives, how long it's retained, and whether learners have meaningfully consented. A transcript of a struggling employee's practice session is sensitive data. Treat it that way.

How do you practice ethical instructional design?

Use a performance consulting mindset: analyze first, recommend the cheapest effective intervention, design practice-based training only when training is warranted, then evaluate and improve. Whatever your job title says, the focus is the solution to the business need, not the course that someone asked you to build.

Analyze the business goal and performance gap
Identify the measurable business goal, the behaviors that drive it, and what's actually preventing people from performing. This is where most course-mill projects die, and should.
Ideate and recommend solutions, training or not
Match solutions to causes. Process fixes, incentives, job aids, and tooling changes all belong on the table alongside training.
Design practice for high-priority tasks
If training makes the cut, build it around practice activities for the highest-priority on-the-job tasks, not around content coverage.
Evaluate with all the data you can get
Use performance data, behavioral observation, and learner feedback to find out whether the intervention moved the needle.
Improve the intervention
Adjust based on what the data shows. Ethical design is a loop, not a launch.

If you're new to this way of working, I've recommended the same starting point for years: action mapping. It's the most approachable on-ramp into performance consulting and up-front analysis that I know of. Ethical practice also includes accessibility. If learners with disabilities can't perceive or operate your course, it fails them by design, so build to accessibility standards from the start rather than retrofitting.

In a 2025 live workshop, I walked through this exact front-end work: tying a project to a measurable business goal, writing a training goal the way Cathy Moore suggests, and only then choosing instructional strategies and evaluation methods. If you'd rather see it done than read about it:

There's a self-interested case here too, and I'll happily make it. Working this way is simply more satisfying. Personally, I'd much rather look back on a project and see that we figured out what people needed to do, why they weren't doing it, and found a solution, even when that solution wasn't training at all, than know I shipped another course into the void.

Should new instructional designers push back too?

Not on day one. Build the core design and development skills that get you hired first, then grow into performance consulting once you're inside and can see whether your courses actually change anything.

This is where ethics advice usually goes wrong: it demands that the person with the least leverage take the biggest stand. My advice since 2021 has been more practical. If your main goal is landing a job that pays well, focus first on core instructional design and development skills, things like writing, visual design, and Storyline. That gets your foot in the door. Then, once you're in and you start noticing that your courses don't seem to be making much of an impact, that's your cue to start bringing in performance consulting approaches. That sequencing is also how I structure the programs at my career school, Peck Academy: learn foundational ID theory, business acumen, and Storyline 360 first, with AI skills layered in.

Even as a junior designer, though, you can practice ethics-lite from week one. You don't have to refuse a course order to ask, "What should people be able to do after this, and what's stopping them now?" Good analysis questions make you look thoughtful, not difficult.

Frequently asked questions

Is it unethical to create a course just because a client asks for one?

Building a requested course isn't automatically unethical; skipping the analysis is. If you've asked why, examined what's causing the performance gap, and training genuinely addresses it, build away. The ethical failure is treating "they asked for it" as a substitute for finding out whether it can work.

What should I do if analysis shows training isn't the best solution but I'm told to build it anyway?

Document your findings, name the specific risk of proceeding, and then build the best version you can. Pair the course with practice activities and performance support like job aids, and schedule a follow-up evaluation so the results speak for themselves. Registering your concerns and then doing excellent work within the constraint is ethical; silent compliance is not.

Is there a code of ethics instructional designers can actually sign onto?

Yes. AECT publishes a Code of Professional Ethics, ISPI publishes a Code of Ethics alongside its performance standards, and ATD publishes a Code of Ethics for talent development professionals. Joining any of them binds you to a code, but no code is enforced across the "instructional designer" title itself, so your day-to-day process matters more than your membership card.

Where does this leave us?

Codes of ethics for our field exist. Enforcement doesn't. Until that changes, the ethics of instructional design live in your process: analyze before you build, recommend the cheapest intervention that will actually work, and design training around practice when training is truly the answer.

And then close the loop. Evaluation is the ethical feedback loop of this whole profession: it's how you find out whether you helped anyone, and it's the evidence that earns you the right to say "no course needed" next time. Turn the lights on before you count your score.