How we use AI on your project

We advise businesses on responsible AI, so it would be strange not to be clear about our own. Here’s where AI helps on your project, where it doesn’t, and what happens to your information.

Last updated: 21 October 2026

Why we publish this

Nobody asked us to.

But we advise clients on AI governance, and we hold ISO 42001:2023, the international standard for AI management. A firm that does both should be willing to say how it uses AI on the work you’re paying for.

It’s also what we’d tell you to do. Being clear about where AI sits in your work is good practice, and increasingly what the EU AI Act expects.

Where AI Helps On Your Project

Research. Gathering and summarising source material, regulations, market information and technical documentation.

Drafting. First versions of documents, specifications and reports, which a person then works on properly.

Code. Writing and reviewing code, generating test cases, and checking for common problems.

Analysis. Working through data to find patterns worth investigating.

In every case, AI shortens the groundwork. It doesn’t do the thinking.

Where We Don’t Use It

The Recommendation.

When we tell you where AI would pay back in your business, that judgement comes from people who have delivered these projects before. It isn’t generated.

Risk Classification.

If we’re assessing your obligations under the EU AI Act, a person makes that call, because getting it wrong has consequences for you.

Architecture Decisions.

What to build, how to build it, and what to leave alone. People decide that.

Anything With Our Name On It.

Every proposal, report and recommendation is read by a person before it reaches you. Nothing goes out unreviewed.

If we delivered something and it’s wrong, that’s on us. Not on a tool.

What Happens To Your Information

This is the part most clients actually want to know.

Your Data Stays In The EU.

Client information is processed within the EU and EEA.

We Don’t Train Public Models On It.

Nothing you share is used to improve anyone’s AI model.

We Have A Policy, And We Follow It.

Client data doesn’t go into public AI tools. That’s an internal rule our team works to, not an aspiration.

We Use Business Accounts, Not Consumer Tools.

Where AI is used on your project, it runs through business accounts configured so your information isn’t retained for training.

Sensitive Material Is Handled Separately.

Personal data, commercially sensitive documents and anything covered by an NDA follow stricter handling than general research.

We’re ISO 27001:2022 Certified

for information security, independently audited.

What We Build For You Works The Same Way

The principles above aren’t just about how we work. They’re how we build.

A Person Approves What Matters.

Every system we deliver has a named human signing off anything that leaves your business or affects a customer. The AI reads, matches and drafts. A person decides.

Everything Is Logged.

You can see what the system did and why, which is what you’d need if a regulator or a client ever asked.

You Own It.

After final payment, the software, the models and the data are yours.

Your Data Stays Where It Should.

EU processing, GDPR designed into the architecture rather than added at the end.

What We Won’t Do

We Won’t Pretend AI Is A Person.

If you’re ever dealing with an automated system from us, you’ll be told.

We Won’t Let AI Make A Decision About You.

No AI system of ours decides whether to work with you, what to charge you, or how to prioritise your enquiry.

We Won’t Use Your Project As An Example Without Asking.

Case studies, testimonials and client names are published only with permission.

We Won’t Put Your Data Somewhere You Wouldn’t Expect.

If a project needs a tool we haven’t used before, we’ll tell you before anything of yours goes into it.

Ask us anything about how we work

Responsibility for how AI is used in our work sits with our leadership team, not with any tool or vendor. If you want to know exactly how it would be used on your project, ask. We’ll answer properly, and if you think we’ve got something wrong, we’d want to hear that too.

Name

Phone number

Tell us a bit about your project or goals

We respond within 24 hours · Your data stays private and GDPR compliant.

Questions clients ask about this

AI Consulting Frequently Asked Questions

AI helps with the groundwork: research, first drafts, code, data analysis. The thinking, the recommendations and the decisions come from people who have delivered these projects before. Everything that reaches you has been read and approved by a named person.

No. Client data doesn’t go into public AI tools, that’s an internal policy our team works to, and where AI is used on your project it runs through business accounts configured so your information isn’t retained for training.

Ask us and we’ll tell you, specifically for your project. We don’t publish a standing list because the tools change, and what’s used depends on what we’re building for you.

Within the EU and EEA. We’re ISO 27001:2022 certified for information security, independently audited.

Yes, and clients in regulated sectors often do. Raise it on the first call and we’ll agree what’s appropriate for your project, including where any constraint affects timelines or cost.

Yes. If a project needs a tool we haven’t used before, you’ll hear about it before anything of yours goes into it.

Yes. Every system we deliver has a named person approving anything that leaves your business or affects a customer, with everything logged so you can see what happened and why.

We are. A person reviews everything before it reaches you, so if something we delivered is wrong, that’s our responsibility, not a tool’s.

ISO 42001:2023 is the international standard for AI management systems. Ours is certified to it and independently audited, which means the oversight, logging and accountability described on this page are externally checked rather than self-declared.

How AI will be used on your project forms part of the specification document we agree before work begins, alongside the quote and timeline. We also sign an NDA before any detailed discussion.