Ask Devin

I think about this pretty simply: AI is great at helping with volume, speed, and structure, but internal comms is still very much about judgment, context, and trust. So the work that benefits most from AI is usually the work that supports thinking — not the thinking itself.

One of my favorite ways to use AI right now is taking the Zoom transcript from our weekly All Hands and using it to help draft the recap email. I still have to tell AI what the real priorities were, add context, and shape the message so it actually makes sense for our employees, but it turns a task that might have taken me an hour or more into something closer to 30 minutes.

The other huge benefit is that it allows me to be more present during those meetings. I’m no longer frantically taking notes or worrying about capturing every detail, because I know I can rely on the transcript later. I think that’s something internal communicators often overlook. Yes, AI can help us get time back, but it can also help us show up better in the moment. You’re an employee, too.

More broadly, I’ve found AI really helpful for drafting first passes, summarizing long conversations or transcripts, pulling out themes from survey results, and creating TL;DRs. It’s also great for editing and clarity checks when you’ve been staring at something too long and just need a second set of eyes. All of that saves time and mental energy without replacing actual decision-making.

AI can also be useful for scaling repetitive work. Things like adapting a core message for different channels or audiences, or brainstorming alternate headlines or subject lines. It’s not making the call for you — it’s giving you options so you’re not starting from a blank page every time.

Where internal comms should always stay human-led is anything that requires nuance, empathy, or a deep understanding of your people and culture. Change communications, leadership messaging, sensitive topics, or moments where trust is fragile really need a human in the driver’s seat. AI can support those messages, but it shouldn’t be deciding what to say or how to say it.

The same goes for relationship-driven work. Partnering with leaders, coaching managers, reading between the lines of employee feedback, and knowing when not to communicate yet — that’s all deeply human work. AI doesn’t know your history, your dynamics, or the emotional context behind a moment.

For me, the line is pretty clear. If the work requires taste, judgment, or a strong point of view, it should be human-led. If the work is about speed, organization, or getting a solid starting point on the page, AI can be a huge help.

Used this way, AI doesn’t replace internal comms. It protects the parts of the job that matter most and gives you back time and presence to do them well.