Our website
Our website describes our consultancy services in general terms. It does not answer questions with AI, and nothing on it is advice for your particular situation. Advice we give you is given under an agreed engagement, on the terms set out in our proposal and our Terms of Service.
The AI systems we help you build
- AI can be wrong. Language models can produce inaccurate, incomplete, inconsistent or out-of-date output, and can state wrong answers confidently.
- Results vary. We test carefully, but we cannot guarantee a particular accuracy, saving or return on investment.
- Keep people in the loop. We recommend testing, monitoring and human review before an AI system is relied on for decisions that matter. Deciding to put a system into live use is yours.
- Third-party models. Systems may rely on models and platforms from providers such as Anthropic or OpenAI. Their terms, behaviour, prices and availability can change.
- Your obligations stay with you. How you use AI in your business must meet your own legal and regulatory duties, including data protection and any rules in your sector.
How we use AI in our own work
We may use AI tools to help with research, drafting and code. A person reviews everything we deliver. We only put a client’s confidential information or personal data into an AI tool when the client has agreed and our data processing agreement covers it.
See clauses 8 and 12 of our Terms of Service.