E3.1Included
Approved tools
Which tools and which accounts people can use, and how to ask for a new one.
Detail E3 — Policy
Simple AI rules your team will follow.
Without rules, some people avoid AI altogether and others paste client details into free tools. A good policy fixes both. We write a short, practical AI use policy for your business: which tools are approved, what data can go into them, what must be checked by a person, and who to ask when unsure.
- Price
- On request
Fig. E3.1 — Before using AI on a taskNot to scale
Which tool?Approved list
What data?P1Allowed · Ask first · Never
Who checks?Before it leaves the business
Table E3 — Scope
What the policy covers.
E3.2Included
Data rules
What can go in, what needs permission and what never does, with examples from your work.
E3.3Included
Checking output
Which work must be checked by a person before it reaches a customer or a decision.
E3.4Included
Being open about AI
When to tell customers AI was involved, and how.
E3.5Included
Responsibilities
Who owns the policy, who approves new tools and who to ask.
E3.6Included
One-page summary
The rules on a single page, for everyone to keep to hand.
Notes — Why it matters
What a policy protects.
- 01Client and personal data, which UK GDPR already expects you to look after, AI or not.
- 02Confidential work, like pricing, contracts and plans, from ending up in tools that keep it.
- 03Your reputation, from AI mistakes reaching customers unchecked.
- 04If you operate in the EU, the AI Act already expects staff who use AI to understand it well enough to use it safely.
Specification — How we write it
Short enough to follow.
- Plain languageWritten for the people using AI, not for lawyers.
- Real examples"Fine", "ask first" and "never" examples taken from your own work.
- Kept currentA review date, because tools and rules change quickly.
General notes — FAQ
Questions about AI policies.
Is this legal advice?
No. We write a practical policy based on how your business works. If you're in a regulated field, we'd suggest your legal adviser reviews it, and we'll work with them.
We're a team of five. Do we really need a policy?
A short one, yes. Five people pasting client details into free tools is still a data problem. A one-page policy takes little time and removes the guesswork.
Does it come with training?
It can. A policy works best when people see what it means for their tasks, which is what our training covers.
Fig. E3.2 — Next step
Need clear AI rules for your team?
Tell us about your team, the tools they use and the data you handle. We usually reply the same working day.