AI in Learning and Development: Uses & Examples

Sep 13, 2024

AI in Learning and Development: Practical Uses, Examples and Risks

Written by: CIPD Help Editorial Team
Last updated: 13 September 2026

AI is already being used in learning and development, but not every use of it makes sense.

It can save an L&D team hours of routine work, help uncover skills gaps and give employees better opportunities to practise. It can also get things wrong.

The useful question is therefore not whether an organisation should “use AI”. It is where AI genuinely improves learning and where a person still needs to make the decision.

Below are five practical examples of how that can work.

What Does AI in Learning and Development Mean?

AI in learning and development means using artificial intelligence to support how workplace learning is planned, created, delivered or evaluated.

For example, an organisation might use AI to analyse assessment results and find an area where employees repeatedly struggle. An L&D practitioner could then use AI to draft practice activities for that particular skill before checking and adapting them for the organisation.

This is different from simply putting a training course online.

A learning management system can store and deliver training. AI can go further by analysing information, generating material, suggesting learning activities or identifying patterns in employee progress.

The technology itself is not the important part. The learning problem it helps to solve is.

1. Finding Skills Gaps More Quickly

Before creating training, an organisation needs to know what people actually need to learn.

AI can help examine information such as:

• assessment results;
• recurring employee questions;
• competency assessments;
• common mistakes;
• performance trends; and
• role or skills requirements.

Imagine a customer service team has recently introduced a new complaints process.

Managers notice that mistakes are increasing, but they do not know whether the problem is product knowledge, communication skills or poor understanding of the new procedure.

Instead of immediately giving everyone another training course, the organisation could analyse existing data to find where mistakes occur most often.

That gives the L&D team somewhere useful to start.

But there is an important limitation.

AI can identify a pattern. It cannot automatically prove why that pattern exists.

The real problem might be poor training. It could also be unclear instructions, a difficult system, unrealistic workloads or inconsistent management.

An L&D professional still has to investigate the cause before deciding that training is the answer.

2. Creating Learning Material Without Starting From Zero

Content creation is one of the most practical uses of generative AI in L&D.

An L&D practitioner could use it to produce a first draft of:

• workplace scenarios;
• knowledge-check questions;
• role-play exercises;
• summaries;
• explanations;
• revision activities; or
• examples for different job roles.

Suppose a company has changed its data-handling procedure.

Instead of writing every learning activity from scratch, an L&D professional could use an approved AI system to draft separate scenarios for customer service employees, managers and new starters.

That can save time.

It does not remove the need for checking.

Generative AI can produce information that sounds convincing even when it is inaccurate, incomplete or based on assumptions. It may also invent company procedures that were never provided to it.

The finished training should therefore be checked against the organisation`s actual policy and reviewed by somebody who understands the subject.

AI can speed up the first draft. A person still needs to decide whether that draft is right.

3. Making Learning More Relevant to Each Employee

Traditional training often gives everybody the same content even when their knowledge, experience and responsibilities are very different.

AI can support a more targeted approach.

Consider two employees completing management development.

One is a newly promoted supervisor who struggles with giving feedback. The other has managed people for several years but needs more help with strategic decision-making.

Giving both employees the same learning journey would make little sense.

Assessment results, previous learning and progress could instead be used to recommend different activities.

The first manager might receive additional practice around difficult conversations. The second might move more quickly through that material and spend more time on strategic scenarios.

This is a better basis for personalisation than trying to put employees into fixed “learning styles”.

What matters is:

• what the employee already knows;
• what the role requires;
• where the employee is struggling; and
• whether the person is making progress.

Organisations should also be careful about what employee information is used to make these recommendations. Personalisation should not mean collecting unnecessary data or allowing an automated system to make important decisions without human oversight.

4. Giving Employees More Opportunities to Practise

Knowing the answer to a question is not the same as being able to handle the situation at work.

AI-supported simulations and conversational tools can give employees more opportunities to practise.

For example, a newly promoted manager could practise explaining a change in responsibilities to a simulated employee.

The simulated employee might respond with:

• confusion;
• disagreement;
• concern about workload; or
• questions about why the change is necessary.

The manager can then try different ways of handling the conversation.

This type of practice can also be useful for:

• customer complaints;
• sales conversations;
• interviews;
• negotiation;
• difficult management discussions; and
• procedural training.

AI-generated feedback can identify a missed point or ask an employee to reconsider an answer.

But it should not be treated as unquestionable expert advice.

Areas involving employment relations, safeguarding, health and safety, legal requirements or other sensitive decisions may require experienced human review.

The value of AI here is more opportunities to practise, not replacing the professional who understands the real situation.

5. Checking Whether Learning Actually Worked

Course completion is one of the easiest L&D measures to collect.

It is also one of the weakest if used on its own.

Knowing that 95% of employees completed a course does not tell an organisation whether they understood it or whether anything changed afterwards.

AI can help L&D teams examine several sources of information together, such as:

• assessment scores;
• common incorrect answers;
• employee feedback;
• learning activity;
• repeated areas of difficulty; and
• relevant workplace measures.

Suppose employees repeatedly answer one scenario incorrectly after completing a training programme.

That gives the L&D team something specific to investigate.

Perhaps the explanation is unclear. Perhaps the scenario is unrealistic. Or perhaps employees understand the training but cannot apply it because the workplace process itself is confusing.

Again, AI can help find the pattern.

People still have to decide what the pattern means.

A Practical Example: Using AI Across the L&D Process

Consider an organisation introducing a new customer complaint procedure.

Before creating training, the L&D team reviews previous complaint cases and speaks to managers to understand where problems currently occur.

AI helps group recurring mistakes and highlights several possible knowledge gaps.

The L&D practitioner then checks those findings rather than automatically treating every pattern as a training problem.

After confirming the actual learning needs, AI is used to draft several short workplace scenarios and practice questions.

A subject expert checks them against the approved complaint procedure.

Employees then receive practice activities based on their role and existing knowledge.

After training, the L&D team reviews assessment results and employee feedback.

One part of the procedure is still being misunderstood.

Instead of running the entire course again, the team changes that particular activity and gives employees another opportunity to practise it.

That is a much more useful application of AI than simply typing:

“Create an employee training course.”

The technology supports the learning process. It does not own the process.

Where AI Helps — and Where People Are Still Needed

The aim should not be to automate as much L&D work as possible.

It should be to use AI where it improves a learning decision.

What Are the Risks of AI in Learning and Development?

The main risk is not simply that AI might give a wrong answer.

Organisations may also need to think about:

• what employee information is entered into AI systems;
• whether that information is actually necessary;
• where information is stored;
• whether automated recommendations could disadvantage particular employees;
• whether employees know when AI is influencing a process; and
• who is responsible for checking important decisions.

The CIPD`s guidance on responsible AI use places emphasis on responsible adoption and the role people professionals have in how AI is introduced and governed.

UK organisations also need to consider data protection when AI systems process personal information. The Information Commissioner`s Office guidance on AI and data protection covers areas including fairness, transparency, accuracy, security and data minimisation.

For L&D teams, a sensible approach is straightforward:

Use approved systems, share only necessary information, check important outputs and keep people responsible for significant decisions.

AI Is Also Creating a New Learning Need

AI is changing more than the way organisations produce training.

It is changing what employees need to learn.

Knowing how to type a prompt is only a small part of AI capability.

Employees may also need to understand:

• when an AI answer should be checked;
• how to recognise unreliable output;
• what information should not be entered into an AI system;
• where human judgement is still required;
• how AI affects their particular role; and
• how to use AI responsibly at work.

The UK Government`s AI Foundation Skills for Work benchmark separates workplace AI capability into technical, non-technical and responsible or ethical skills.

That gives L&D teams two different questions to answer:

How can AI improve the way we deliver learning?

and

What do employees now need to learn because AI is changing their work?

A useful L&D strategy needs to consider both.

What AI in L&D Means for CIPD Learners

AI can appear in CIPD work through several different topics rather than only through a question specifically about artificial intelligence.

It may be relevant when discussing:

• learning and development;
• workforce skills;
• organisational performance;
• digital working;
• people strategy;
• employee experience;
• ethical decision-making; or
• organisational change.

Simply listing AI tools will normally produce a weak discussion.

A stronger approach is to start with the organisational problem.

For example, instead of only writing:

“AI can personalise employee learning.”

Ask:

• Why is personalisation needed?
• What employee information would be used?
• What would AI actually change?
• What benefit should that produce?
• How would the organisation know whether it worked?
• Could the approach create privacy, fairness or accuracy problems?

That moves the discussion from description to analysis.

Students who are struggling to interpret an assessment criterion or build that level of analysis can use our CIPD assignment help UK service for support with Levels 3, 5 and 7.

Frequently Asked Questions

How is AI used in learning and development?

AI can be used to identify possible skills gaps, create learning material, personalise training, provide practice activities and analyse learning results. Its use should begin with a genuine learning need rather than simply introducing AI because the technology is available.

Can AI identify employee skills gaps?

AI can help identify patterns in assessment, performance or workforce data. However, finding a pattern does not prove that lack of training caused the problem. L&D professionals should investigate the cause before recommending training.

Can AI replace learning and development professionals?

AI can automate or speed up parts of L&D work, particularly analysis and content drafting. It cannot replace the organisational understanding, professional judgement and accountability required to decide what employees actually need.

How can AI personalise employee training?

AI can use suitable information about existing knowledge, assessment performance and progress to recommend different learning activities. Organisations should still consider fairness, transparency and how employee information is being used.

What are the risks of using AI in L&D?

Risks can include inaccurate information, bias, inappropriate use of employee data, weak transparency and excessive reliance on automated recommendations.

How should an organisation introduce AI into L&D?

Start with a specific problem rather than selecting an AI tool first. Decide what needs to improve, test AI on a limited use case, keep human review in the process and agree how the result will be measured.

Is generative AI useful for creating workplace training?

Yes. It can be useful for producing first drafts of scenarios, examples, quizzes and explanations. However, the material should be reviewed for accuracy, context and relevance before employees use it.

Final Thought

AI does not improve workplace learning simply because it makes training quicker to produce.

Its real value appears when it helps an organisation understand a learning need more clearly, gives employees better opportunities to practise or provides information that helps improve the next learning decision.

For an L&D professional, the useful question is therefore not:

“Where can we add AI?”

It is:

“Which learning problem can AI help us solve better without removing the judgement, context and responsibility that people still need to provide?”