WAJD Learning

Module 1 of 2 · 45 minutes

The same rules for an AI agent: responsibility, honesty and training

By the end of this module you will be able to

  • State who is responsible for an AI agent
  • State the penalties
  • Decide when to tell customers
  • List what the agent must respect
  • Explain why testing comes first

Work through it

1 interactive for this module, built on the WAJD Teach engine. Nothing moves until you ask it to, and every one has a written version if you would rather read it.

Watch: Emma and George talk it through

2 minutes. Captions are on, and the same conversation is written out in full below. The voices are computer generated.

Emma George, we've put an AI agent on our shop's chat. If it tells a customer something wrong, isn't that the supplier's problem?

George No. The Competition and Markets Authority published guidance on 9 March 2026. It says you're responsible for what an AI agent does in the same way you're responsible for what an employee does.

Emma Even though we didn't build it?

George Even if someone else designed or provides the agent on your behalf.

Emma What's the headline?

George One sentence. When it comes to dealing with your customers, the same rules apply whether using AI or human agents.

Emma And if we get it wrong?

George You could be fined up to ten per cent of your worldwide turnover. And possibly forced to compensate the customers affected.

Emma Do we have to say it's AI?

George The test is this. If the fact they're dealing with AI rather than a person might affect people's decisions, then you should tell them. Don't let customers think it's a real person.

Emma Our marketing team wants to call it our AI expert.

George Careful. The guidance also says don't overstate the role of AI, or what it can or can't do.

Emma What does the agent need to know?

George Three things. To respect customers' statutory rights and the terms of contracts, like cancellation rights. To avoid misleading customers, by what it says and what it doesn't say. And to properly obtain any consents the law requires.

Emma It's had none of that. We just switched it on.

George Then it's guessing. An agent that's never been given your returns policy and the law on faulty goods will make both up.

Emma What should we have done first?

George Tested it. The guidance calls testing a crucial part of training. You use the results to check and improve it before you deploy.

Emma And the supplier?

George Ask how it was trained on consumer law and what the tests showed. If someone else provides the AI and you don't check they've done that, you may break the law.

The written material

The same rules apply

The Competition and Markets Authority published Complying with consumer law when using AI agents on 9 March 2026. It is encouraging about AI, and plain about the law: consumer law requires you to treat your customers fairly, and it does not matter whether they interact with a person or an AI agent.

You are responsible for what an AI agent does in the same way you are responsible for what an employee does. That is true even if someone else designed or provides the agent on your behalf.

Tell your customers

Under consumer law you should give consumers the information they need to make informed decisions, and you should not mislead them. The guidance says to consider whether you need to label an AI agent so that customers are not misled into thinking a service is being provided by a real person.

The test it gives: if the fact they are dealing with AI rather than a person might affect people's decisions, then you should tell them. It adds a second point that is easy to miss. Do not overstate the role of AI in providing a service, or what it can or cannot do.

Train the agent on the law

Think about what the agent will be set up to do and how that might affect customers, directly or indirectly. Then think about the data it needs and how it will be prompted to do three things.

  • Respect customers' statutory rights and the terms of contracts, such as cancellation rights
  • Avoid misleading customers, through what it says and what it does not say
  • Properly obtain any consents that consumer law requires

Test before it meets a customer

The guidance calls testing a crucial part of training AI agents, and says to use the results to check and if necessary improve the agent's performance before you deploy it.

It also closes the obvious gap. If you do not take appropriate steps to train AI agents, or if someone else is providing the AI and you do not check that they have done so, you may break the law. So when you buy an agent from a supplier, ask how it was trained on consumer law and what the tests showed. The guidance suggests using it as the basis for that conversation.

Knowledge check

The knowledge check and your certificate need a free account, so that your progress and results can be saved as evidence.

Create a free account Sign in

The learning itself stays free and open. You are reading all of it right now without an account.