Guide

AI policy for IT service providers and MSPs

An AI policy for IT firms and managed service providers: client credentials, code and logs in AI tools, and how to answer clients who ask about your AI rules.

Updated 2026-11-19

IT firms use AI for code, scripts, documentation and support. You also hold the keys to your clients' systems, which makes your AI rules part of your clients' security.

Where AI helps in an IT firm - Writing and reviewing scripts and code - Explaining errors and logs - Drafting documentation, tickets and client updates - Summarising vendor documentation

The specific risks - Secrets in prompts. Passwords, API keys and config files pasted into AI tools. - Client data in logs. Logs and tickets often contain personal data and internal details. - Unreviewed code. AI-generated scripts that work, but open a security hole.

Rules to put in your AI policy 1. Only approved tools, on business accounts with training on your data switched off. 2. Never paste credentials, keys, tokens or client config into AI tools. Redact first. 3. Strip personal data and client names from logs before asking for help. 4. AI-generated code is reviewed like any other code before it reaches a client system. 5. Follow each client's rules on AI use; many now ask in security questionnaires. 6. A named person approves new AI tools and keeps the list.

In the EU Article 4 of the EU AI Act asks companies that use AI tools to ensure sufficient AI literacy among their staff. A written policy, a tool register and a documented training are also exactly what clients' security questionnaires ask for.

Our kit writes this policy with your firm's name and tools, and trains your team in 15 minutes with a certificate per person. Your clients can use it too: see our partner programme.

Make your AI policy in 5 minutes

The free generator writes a policy with your company name and tools. The full kit adds the tool register, the training plan and a 15-minute staff training with certificates.

More guides