Reading · 9 min · Lesson 2 of 6
What a good policy covers
Five elements: tools, data, review, labelling and accountability.
Approved tools: which AI tools are allowed and for what, including whether company or personal accounts may be used. Data rules: what must never be pasted into a public AI tool, such as client data, personal data, trade secrets and source code.
Human review: which outputs always need a person to check them before they are published or sent outside. Labelling: when AI-generated content must be marked, in line with the EU AI Act and client expectations. Accountability: the person who uses AI is responsible for the result, including its mistakes.
Write each element as a rule with an example. “Do not paste client data into unapproved tools” is a good rule; an example such as a client contract or a list of customer email addresses makes it unmistakable.
Key takeaways
- Cover tools, data, review, labelling and accountability.
- The person who uses AI is responsible for the result.
- Pair every rule with a concrete example.