AI Communication Templates You’ll Need Before Something Goes Wrong
Your "Practical Friday" free download!
Most AI governance frameworks fall apart not because the policy was wrong, but because nobody had prepared what to say when the policy needed to be communicated. A system produces biased output, and a manager improvises something in a Slack message that legal then spends two weeks walking back. A vendor gets a vague complaint instead of a specific performance notice, responds to the vagueness accordingly, and nothing changes. A partner finds out about an AI incident from a customer rather than from the account team.
The templates in this week’s download come from my forthcoming book Guardrails: How To Embrace AI in Your Business Without Damaging Your Brand, out in October. They cover five categories:
Customer AI disclosure language for four scenarios: AI-assisted content (a human reviewed it), AI-generated content (AI produced it with post-hoc review), AI-powered features, and the chatbot handoff that tells a customer they’re talking to a machine and how to reach a person.
Crisis statement frameworks for the first 24 hours when you don’t have all the answers yet; the complete response once the investigation is done; a bias incident statement for when a system has produced disparate outcomes across groups; and the privacy breach notification format that covers the legal basics without making things worse.
Employee policy communication for introducing AI governance without sounding like a surveillance operation, the vendor approval request that explains the process without being defensive, and the incident reporting template that makes it safe for people to flag problems early rather than hoping someone else notices.
Vendor communication for due diligence requests that are treated as standard processes rather than accusations, and a performance issue notice that is specific enough to elicit a real response.
Partner communication for the AI disclosure notice you should be sending before AI touches a shared workflow, and the incident notification that preserves a partnership even when the news is bad.
Each template includes a brief note on what makes the specific language work, since adapting a framework you understand produces better results than filling in blanks you’re guessing at.
Ethicore Advisors Author’s Note
Craig McDonogh is the founder of Ethicore Advisors and the author of the forthcoming book “Guardrails: How To Embrace AI in Your Business Without Damaging Your Brand,” available October 2026.