- 1: Part 1, Understanding AI agents
- 2: Part 2, The six principles
- 3: Part 3, Introducing an agentPart 3, Introducing an agent
- A–E: Appendices
- 3.1: Preparing to introduce an agent
- 3.2: Putting changes into practice
- 3.3: The opening hours example concluded
Part 1, Understanding AI agents
Part 3, Introducing an agent
Chapter 3.1
Preparing to introduce an agent
Scope and limitations of this guide
This guide provides general information on organising work with AI agents, not advice tailored to a particular organisation, system or decision. It is not legal or regulatory advice.
Its examples, checklists and templates do not establish that a proposed use is safe, lawful or suitable. You need to assess and test the proposed use in your own circumstances. Following the guide does not by itself meet the requirements of any law or standard.
The guide is provided in good faith, using information available at the time of publication or revision. AI systems, provider terms and legal requirements can change. Before acting, check the current requirements that apply to your organisation and country, confirm what the system can do and test the proposed use. You are responsible for how you apply the guide and should obtain professional advice where appropriate.
Reading, downloading or using the guide does not by itself create an adviser–client relationship with Tom Clowry or Considered Ideas Ltd.
To the fullest extent permitted by law, Tom Clowry and Considered Ideas Ltd accept no liability for loss or damage arising from use of, or reliance on, this guide. Nothing in this notice excludes or limits liability for death or personal injury caused by negligence, fraud or fraudulent misrepresentation, or any other liability that cannot lawfully be excluded or limited.
In this chapter
Before an agent becomes part of ordinary work, check that the AI service you have chosen supports the permissions, checks and records you have planned, and decide who needs to be told.
Questions to ask about an AI service or its supplier
Can you limit what it can reach?
Separate accounts, read-only access and permissions for each connection.
Can you require approval?
For the actions you have decided need it.
Does it keep a record?
Of what the agent did, which tool it used and what changed.
Further questions to ask#
What can be stopped, cancelled or reversed?
Check separately for queued, running and completed actions.
What happens to your information?
Ask how the service stores, uses, shares and deletes it, and whether the provider uses it to train its models, as Principle 2 explains.
How is it charged?
Charges that rise with use can grow quickly as an agent takes on more work.
Can you change supplier?
Check whether you can export your instructions, records and saved notes.
Ask the people responsible for information security and contracts in your organisation to assess the supplier as well.
Which AI services colleagues can use#
Colleagues may already use AI services in their work, including services that the organisation has not approved. The organisation needs to decide which services can be used, for what and with which information. That decision needs to involve the people responsible for information security, data protection and contracts. Colleagues need to be able to find the list of approved services easily. Services not on the list must not be used for work, and no information can be supplied to any service unless the list permits it.
Approved services
Which services can be used, and with which accounts and settings.
Permitted uses
The kinds of work each service can be used for.
Information
The information approved for each service. No other information can be supplied to it.
Proposing another
How a colleague can propose a new service or a new use.
Review the list when needs change, and ask colleagues what they need.
Other legal questions#
This section is not legal advice. Its examples are drawn from UK law at the time of writing. The law can change, and the rules may differ in the EU and elsewhere. Ask the person who advises your organisation which rules apply to you.
Data protection and equality law are not the only legal questions. Before you use an agent, ask the person who advises your organisation about each of the following.
What you supply
Do not supply documents, data or other people’s work to an AI service unless the person who advises your organisation has confirmed that you are permitted to.
What the agent produces
Who owns the result, and whether you can publish or sell it.
Contracts
Your contracts with customers and suppliers may limit how information is used, or require you to tell the other party.
Customers’ rights
Using an agent does not remove your obligations to customers. Consumer law can still apply when a customer deals with an agent rather than a person. In the UK, for example, the Competition and Markets Authority has said that a business is responsible for what an agent does, as it is for what an employee does, even if another company supplies the agent. Check how customers can exercise their rights, including any rights to cancellation, refunds or redress.
Sector rules
Professional and sector regulators may publish their own guidance on AI. Check whether yours does.
This guide is not legal advice.
Source CMA, Complying with consumer law when using AI agents, 9 March 2026, updated 30 October 2026.
Telling people about the agent#
Tell people when they are dealing with an agent or when its role materially affects a decision or service they receive. Explain how to contact a person and how their information is used. Colleagues also need to understand the agent’s role, who is responsible and how to raise concerns.
Check that your privacy information covers the use of AI before the agent is introduced, as Principle 2 explains.
© 2026 Considered Ideas Ltd. Licensed under the Creative Commons Attribution 4.0 International licence (CC BY 4.0). This guide provides general information and is not legal or regulatory advice. The guide’s home page gives the suggested credit.
Scope and limitations of this guide
This guide provides general information on organising work with AI agents, not advice tailored to a particular organisation, system or decision. It is not legal or regulatory advice.
Its examples, checklists and templates do not establish that a proposed use is safe, lawful or suitable. You need to assess and test the proposed use in your own circumstances. Following the guide does not by itself meet the requirements of any law or standard.
The guide is provided in good faith, using information available at the time of publication or revision. AI systems, provider terms and legal requirements can change. Before acting, check the current requirements that apply to your organisation and country, confirm what the system can do and test the proposed use. You are responsible for how you apply the guide and should obtain professional advice where appropriate.
Reading, downloading or using the guide does not by itself create an adviser–client relationship with Tom Clowry or Considered Ideas Ltd.
To the fullest extent permitted by law, Tom Clowry and Considered Ideas Ltd accept no liability for loss or damage arising from use of, or reliance on, this guide. Nothing in this notice excludes or limits liability for death or personal injury caused by negligence, fraud or fraudulent misrepresentation, or any other liability that cannot lawfully be excluded or limited.

