Guidance

Complying with consumer law when using AI agents

Published 9 March 2026

The UK is the third largest AI market in the world, after the US and China, and is ranked number one in Europe and fourth in the world for AI, according to the Global AI Index.

AI has the potential to boost economic growth and improve people’s everyday lives, and the Competition and Markets Authority (CMA) is committed to encouraging its use.

Your business might already be exploring ways to use agentic AI innovatively. For example, using AI agents to:

• handle customer queries
• process refunds
• recommend products
• manage marketing campaigns

Using AI agents can help make your business model more productive and efficient. It can also improve your customers’ experience. But to benefit fully from agentic AI, you need to use it responsibly and in compliance with consumer law.

If your customers do not trust or have confidence in your AI or AI-driven services, they are less likely to use them and benefit from them, and any innovative and disruptive use of AI may not reach its full potential.

This page provides guidance on how your business can use agentic AI to engage with its customers, while staying on the right side of the law and doing right by them.

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

Consumer law requires you to treat your customers fairly. It does not matter whether they interact with (or get information produced by) a person or an AI agent. It’s important to remember that you are responsible for what an AI agent does in the same way you are responsible for what an employee does. This is true even if someone else designed or provides the AI agent on your behalf – so make sure you think about compliance with consumer law if you are procuring something from a third party, for example using this guidance to discuss with them.

If you break consumer protection law, you could face action by enforcement authorities (including the CMA). You could be fined up to 10% of your worldwide turnover, and possibly forced to compensate affected consumers.

Find out more about how we use our consumer protection enforcement powers and direct consumer enforcement powers.

Tell your customers if you use an AI agent

Businesses may use AI agents in different ways, which can be confusing for people. Being clear and open about how you use AI agents is a good way to build trust, especially if it might be a surprise. Under consumer law, you should give consumers the information they need to make informed decisions, and you should not mislead them.

If you use an AI agent, consider whether you need to label it so you do not mislead customers into thinking that a service is being provided by a real person – if the fact they are dealing with AI rather than a person might affect people’s decisions then you should tell them. Do not overstate the role or AI involved in providing a service, or what it can or cannot do.

Train your AI agents to comply with consumer law

Think about what the AI agent will be set up to do, and how that might affect your customers, directly or indirectly.

AI agents often require relevant data to perform specific tasks well. Think about what data your AI agent will need, and how it will be prompted to:

• respect customers’ statutory rights and the terms of contracts (for example to make sure cancellation rights aren’t being breached)
• avoid misleading customers (both through what they do and do not say)
• properly obtain any necessary consents required by consumer law

Testing is a crucial part of training AI agents; this includes evaluating performance, for example through A/B testing or unit testing. Use the results of any testing to check and if necessary improve the AI agent’s performance before you deploy it.

If you do not take appropriate steps to train AI agents – or if someone else is providing the AI, and you don’t check that they have done so – you may break the law. Ultimately, you will be responsible if an AI agent does something illegal, so it is important to make sure you think about compliance with consumer law from the start.

Monitor how your AI agents are performing

Regularly check how well an AI agent is performing. Make sure it is:

• delivering the right results
• behaving as intended
• complying with consumer law

This will help you to spot potential problems.

Some AI models can misinterpret data and ‘hallucinate’ results that are nonsensical or inaccurate. Make sure there is a human in the loop, actively checking that the AI agent is making correct decisions and generating expected results. Regular human oversight is important to catch mistakes and ensure that AI agents are completing tasks in a legally compliant way.

To make sure you comply with your legal obligations, it is important that you have appropriate processes in place to monitor AI agents.

Refine the AI agent quickly if there’s a problem

If you realise that an AI agent is not performing as expected, and that this is leading to non-compliant or potentially non-compliant outcomes, act quickly to address the problem, for example by refining prompts or workflows. It’s especially important to act quickly if an AI agent interacts with large numbers of people (or its outputs could reach a lot of people) or vulnerable customers.

If you do not act quickly to address problems, you may end up breaking the law – ultimately, if an AI agent does something illegal, you will be responsible.

Example: using an AI agent to run your latest marketing campaign

Think about the requirements of the Digital Markets, Competition and Consumers Act 2024 when you design and train the AI. For example:

• any marketing materials should provide accurate information about prices (which should include all unavoidable charges) and products
• paid endorsements should be properly labelled
• any offers or price reductions should be genuine

Someone with appropriate experience should review any AI-generated marketing campaign regularly, to make sure that the AI agent is behaving as intended. In particular, check any public-facing information or statements made to customers.

If you identify a problem – for example, that the AI agent is not consistently requiring social media influencers to label their endorsements, or making appropriate checks that this has been done – then quickly refine the relevant prompts and workflows.

Example: using an AI agent to process refund requests

Think about the contractual and legal requirements which apply to refunds when designing and training the AI, for example taking into account:

• consumers’ statutory rights under the Consumer Rights Act 2015 (for example, if goods are faulty) and the Consumer Contracts (Information, Cancellation and Additional Charges) Regulations 2013 (for example, if they change their mind)
• any terms and conditions in your contracts with customers (for example, extended returns periods)

Someone with appropriate experience should regularly review exchanges with consumers to make sure that the AI agent is behaving as intended. In particular, check that the AI agent’s decisions properly reflect the nature of a customer’s refund request, and any differences in time limits or amounts due.

If you identify a problem – for example, that the AI agent is not always taking into account your extended returns policy when deciding whether a customer is entitled to return goods and claim a refund – then quickly refine the relevant prompts and workflows.

Example: using an AI agent to respond to customer service queries

Think about the requirements of the Digital Markets, Competition and Consumers Act 2024, and consumers’ statutory and contractual rights, when you design and train the AI (read the examples above).

For example, you must respond accurately to queries about prices, products and rights, give consumers all the information they need to take informed decisions, and not make it difficult for them to exercise their rights.

Someone with appropriate experience should regularly review how the AI agent responds to different queries, and any complaints or feedback from customers, to make sure that it is behaving as intended.

If you identify a problem – for example, that the AI agent keeps giving people the wrong information about the features of a product or their cancellation rights – then quickly refine the relevant prompts and workflows.

Example: using an AI agent to provide a service to customers

You provide a deal comparison service. The service uses an AI agent to ask customers for their preferences, then find and recommend the cheapest matching deals.

Think about the requirements of the Digital Markets, Competition and Consumers Act 2024 when designing and training the AI. For example, results should be accurate, and important information should be clearly and prominently disclosed, such as:

• how much of the market has been looked at
• what data has been searched
• any particular limitations in the results

Search results should also match a customer’s search criteria, and things like the coverage of any search, how results are ranked and any links with suppliers should be transparently disclosed.

Someone with appropriate experience should regularly review how the AI agent responds to different queries, as well as any complaints or feedback from customers, to make sure that is behaving as intended.

If you identify a problem – for example the AI agent keeps including deals which do not match the customer’s search criteria – then quickly refine the relevant prompts and workflows.

Find out more about consumer protection law

The CMA has more guidance on your requirements under consumer protection law, including on:

• treating consumers fairly
• writing fair contracts

Business Companion also provides impartial legal guidance for businesses.

The Digital Regulation Cooperation Forum’s AI and Digital Hub has some helpful case studies.