AI Usage Policy
We think transparency about AI matters. This page sets out where AI is used in the studio's work, what the assistant on this website is and is not, and who stays responsible for what.
What this policy covers
The Creativity Lab is a creative studio: we design and build websites, stores, applications and digital products. This policy covers how we use AI as a tool inside that production, and the assistant on this website. Where an engagement involves building or operating an AI system, the scope, safeguards and responsibilities are agreed in writing per engagement; this policy does not replace those agreements.
AI as a tool in our production
- Drafting, rewriting and translating copy.
- Code assistance during development.
- Exploring image, design and content directions.
- Research and summarising source material.
AI never designs, writes or ships on its own here. Every result is reviewed, edited and approved by a person before it is published or delivered.
The assistant on this website
The Studio Concierge on this website answers questions with a language model. When you send a message, your message, the earlier messages of the conversation, the choices you made in it and the page you are on go to our server, and from there to Anthropic, whose model writes the answer. The answer draws on the pages of this website, and links in an answer are checked against a fixed list before you see them. If answering with the model does not work, a scripted assistant in your browser answers that question instead. An answer can be wrong or incomplete: for a quote, an appointment or anything else that matters, contact us directly, and please do not share confidential information or passwords through the assistant. A copy of the conversation is also kept in your own browser’s session storage and disappears there when you close the tab.
Human review and responsibility
Parts of the content on this website and in our deliverables may have been drafted with AI assistance. AI output can contain errors, omissions or bias. All content is reviewed, edited and approved by us before it is published or delivered. The client remains responsible for checking and approving final deliverables before they go into production.
Ownership of output
Subject to full payment, assets created for a project are delivered to the client as project deliverables, as set out in our intellectual property policy. The underlying models, third-party services and our own reusable tooling remain the property of their respective providers and of us.
Acceptable use
- No unlawful, harmful, deceptive or fraudulent applications.
- No content that infringes the rights or privacy of others.
- No use of AI to unlawfully impersonate, manipulate or track people.
We reserve the right to refuse or stop any request that is unethical or unlawful, or that materially conflicts with our values or applicable law.
Limitations and no warranty
AI technology is powerful but not perfect. It can make mistakes, invent information and present it as fact, or return outdated information. Where an engagement involves personal data being processed by an AI provider, we agree that in writing up front and choose privacy-respecting options where possible. We do not warrant that AI-assisted output is fully accurate, complete or fit for every purpose; review and responsibility stay with people.
