We welcome AI and actively use it in our work. It helps us explore ideas, solve problems and build better digital products, giving our team more room for creativity, critical thinking and the work that matters to our clients.
We support AI when it is used responsibly and ethically. For us, that means being thoughtful about where it helps, honest about how we use it, and accountable for the work we deliver. It also means considering its impact on people and the planet.
Here’s what we commit to.
We use AI to help with research, ideas, analysis, problem-solving and software development. We’re curious about what it can offer and open to trying new ways of working.
We choose tools to suit the job. Sometimes AI helps us get to a better result or saves time we can spend elsewhere. Sometimes a simpler approach makes more sense.
Our team sets the direction, makes the decisions and stands behind the results. A named person is responsible for every piece of AI-assisted work.
We review that work before it is used or shared. In software development, we check every changed file, with automated quality checks providing another layer of review. We also check research and written content before relying on them.
People stay in control of decisions and conversations with clients. We don’t run fully automated, unsupervised AI processes in production.
When AI has played a substantial role in work a client receives, we say so. We’re happy to explain what it helped with and how we checked the results.
We take responsibility for everything we publish. If we use an AI-generated image or other creative asset directly, without substantial human reworking, we label it clearly.
For work involving client or personal information, we only use AI tools with data-handling terms appropriate for business use. We share only what is necessary and permitted for the task, respecting confidentiality, privacy and the purpose for which the information was given to us.
We never paste passwords, credentials, API keys or other secrets into AI chat tools. We don’t allow client data to be used to train third-party AI models without the client’s explicit permission.
AI can reflect biases and assumptions that are easy to miss. We check for these in our work, especially where the results could affect how someone is represented or treated, and address problems we find.
We also consider who benefits from the way we use AI and who might be left out. Fairness and inclusion are part of deciding whether an approach is right for the people it affects.
Responsible use includes thinking about AI’s environmental impact and the practices of the companies behind it.
We favour efficient tools suited to the task and consider whether AI is needed. When choosing providers, we consider their environmental and ethical practices alongside security, data protection and how well their tools meet our needs.
Using AI well means understanding what it can do, where it falls short and when to question its output. We help our team build that understanding and keep it up to date.
We commit to meeting the legal and data-protection requirements that apply to our use of AI. If an AI-related error reaches a client, we correct it and look at what we need to change to reduce the chance of it happening again.
As the tools and our understanding develop, we’ll keep improving how we work.
We review this pledge at least once a year, and sooner if our tools, ways of working or relevant requirements change. We also check whether our review processes are catching problems as intended.
Human review, clear accountability and careful handling of client information are already part of our work. We’re building on those practices with:
We’re optimistic about what AI can help us achieve. We want to make good use of it, with care for the people who trust us and responsibility for the work we do.
