Imagine this scene
You’re drafting a message about a sensitive topic. Speed matters, but accuracy is critical. You could rush to meet deadlines and hope nothing is wrong — or you could use AI to accelerate the process, but only if you manage the risks. AI can be powerful, but it introduces hazards that communicators must manage.
AI can produce content that sounds plausible but is factually incorrect. This is especially dangerous when preparing external communications, press statements, or leadership briefings. Always verify critical facts before using AI-generated content. Never assume AI-generated statistics, quotes, or citations are accurate.
AI reflects the data it was trained on. It may inadvertently reproduce biases or omit perspectives. Communicators must be vigilant to ensure content is inclusive, fair, and aligned with organizational values. Review AI outputs for tone, representation, and assumptions.
Never paste sensitive or confidential information into AI tools without considering security and privacy implications. Think of AI as a junior assistant who doesn’t have discretion — it will generate outputs based on what you feed it. What goes in may inform the model or be visible to others.
The core principle is accountability. AI can suggest, summarize, or reframe, but it cannot take responsibility. You remain the editor, the strategist, and the final authority.
The Mindset Shift
AI is fast and capable, but you are responsible. Speed does not replace judgment.
Set your personal AI guardrails before you need them.
List three types of content you regularly work on: one that's public-facing, one that's sensitive or confidential, and one that involves facts, data, or statistics.
For each type, identify the biggest risk if AI gets it wrong — reputational damage, a data breach, a factual error, or biased framing.
Write one clear personal rule for each risk area. For example: "I will never paste employee data into AI tools" or "I will verify every statistic AI generates before using it."
Think about a recent piece of AI-generated content — yours or someone else's. Can you now spot where a factual error, bias, or unverified claim might have slipped through?
Write your three rules somewhere you'll actually see them — a sticky note, a pinned doc, or the top of your working notes. Rules only work if they're visible.
