The rules, guardrails, and timeline for rolling out AI across the organisation responsibly — from day one.
Without governance, AI adoption stalls or creates risk. These three non-negotiable rules and a clear 30-day plan ensure the team moves fast and stays safe.
Never paste client-confidential data (contracts, personal data, financials of clients) into anything outside the ChatGPT Business/Team workspace — Business/Team accounts don't train on your data by default, but always sanitize identifying details when in doubt.
No fully autonomous client-facing sends — every AI-drafted email/proposal/ticket response to a client gets a human review before it goes out, at least for the first 90 days.
One shared home for prompts and GPTs — every useful prompt goes on the Prompt Wall; every useful Custom GPT gets published to the Team workspace, not kept personal — otherwise you rebuild the same wheel 4 times.
Keep this visible — revisit it at each weekly check-in.
This is now a leadership expectation, not an optional experiment. If leadership models AI-enabled behavior, the rest of the organisation follows.