How we use AI on client projects, with room for your own AI policy
AI is part of how we research, design, prototype, test and build software, and this page sets out where it sits. The tools change from project to project. Our responsibility for the work we agree to deliver stays the same.
This page sets out how we work. The commitments for your project are written into its agreement, where you can hold us to them.
Where AI fits into client work, and when it is optional
Depending on the project, our teams use AI tools to organise research material, explore alternatives and inspect product interfaces. They also use them to develop prototypes, write or review code and support quality checks. People make the product choices and assess the work before it reaches you.
You decide how far AI reaches on your project, from wide use to none at all. Within that, how much we use depends on the task and the project.
AI tools are taking on longer and more independent tasks, and our use of them will grow with that. Two things hold as it does: a person assesses what AI produces, and your rules are agreed before work starts.
AI and client confidentiality: how project information is handled
We work with third-party services, including AI features within design, development and collaboration tools. Providers handle data in different ways, depending on the product and the account, so we choose the tools for your project with your rules in hand.
We treat project information according to the confidentiality, access and data-use terms that apply to your project. If a project restricts which AI tools we use, or what we share with them, we agree those limits with you. We record them in the project agreement.
Who reviews AI-assisted work, and who is responsible for it
A person assesses AI-generated material against the work it supports. Product choices, research conclusions and finished work stay with the people responsible for them, and those people review them. Code intended for release goes through the review, testing and acceptance steps agreed for the project.
The project agreement sets out our responsibilities, your team's responsibilities and any release approvals. Using an AI tool moves none of them.
Tell us your AI and procurement requirements before we start
If your organisation has AI, security, privacy or procurement requirements, these steps come before any work starts:
- You tell us your requirements when we plan the project.
- We discuss suitable tools with you, with the environments and restrictions that go with them. For a particular tool or provider, we can go through its published terms with you.
- We record specific commitments in the project agreement.
- We tell you before work begins if a requirement changes the approach, scope or cost.
Our Privacy Policy covers personal information more generally. To discuss a project, contact us.
Questions about how we use AI on client projects
Which AI tools does Tcules use?
The tools change from project to project, and include AI features inside design, development and collaboration software. When we plan your project, we can agree which tools are used and record that in the project agreement.
Can we ask Tcules to work without AI?
Yes. AI is optional, and you can work with us either way. If the restriction changes the approach, scope or cost, we tell you before work begins.
Do AI providers keep or train on our material?
It depends on the provider, the product and the account, because their data-handling arrangements differ. If your project limits what may be shared with AI tools, we agree those limits with you and record them. For a particular provider, we can go through its published terms with you.
What should a design studio's AI use policy cover?
It should say which tasks use AI, how client information is handled, who reviews what AI produces and how a client's own rules are agreed and recorded. This page covers each of those in turn.