Last updated 2 August 2026

AI Policy

Captivated builds software, and some of it uses artificial intelligence. We also use AI tools in our own work. We think it is right to be open about that: how we use AI, the controls we keep around it, and the lines we choose not to cross. We will keep this current as our practice and the law evolve.

Our principles

  • Human oversight. People stay in charge. AI assists our work; it never replaces the person accountable for it.
  • Privacy and sovereignty. We favour local, on-device, private-by-default approaches, so your data stays yours.
  • Transparency. We are open about where and how AI is involved.
  • Proportionality. The more a decision could affect someone, the more human scrutiny it gets.

Human review, always

  • Anything produced with the help of AI is reviewed by a person before it is used in production or shared with a customer. Nothing ships on an AI's say-so alone.
  • We do not let AI make consequential decisions unattended. Where AI informs a decision that affects a person, a human makes or confirms that decision and remains accountable for it.
  • A named person is responsible for every output. "The AI did it" is not an answer we accept, or one we would offer you.

How we use AI

  • In our products. Some of what we build is itself an AI system. We design these to keep the user in control, private by default, and clear about what the AI is doing.
  • In our own work. We use AI to help us write, code and research, always under the human-review rule above. It is a tool in our hands, not an unattended worker.

Your data

  • We do not use your personal data to train AI models without your consent.
  • Where we can process data on your device or in a private environment rather than sending it to third parties, we prefer to. See our Website Privacy notice for how we handle personal data generally.

Lines we won't cross

  • We will not use AI to deceive you. Where you are interacting with an AI rather than a person, we will make that clear.
  • We will not deploy AI to make automated decisions with legal or similarly significant effects on people without meaningful human involvement.
  • We will not build on, or use AI with, data we are not entitled to use.

Standards and regulation

We design our AI work with the direction of UK and EU regulation in mind, including the risk-based, transparency and human-oversight expectations found in frameworks such as the EU AI Act. As the rules develop, we intend to keep pace and update this policy.

Questions

If you have any questions about how we use AI, email [email protected].