AI

Hi Devin,

Our leadership team is really excited about using AI to “move faster,” but I’m worried about speed coming at the expense of clarity and trust. How should internal comms teams think about introducing AI responsibly without creating more confusion or noise?

I’m so glad you asked this because we just did this at Workshop. And honestly, the biggest reason it went well is because it was a truly collaborative effort between the leadership team, department heads, and comms. We aligned early on what AI should be used for, what it definitely shouldn’t be used for, and how fast we actually wanted to move — not just how fast we could move. Not all outputs are good ones!

We did our official launch during an all-company retreat, which gave us the space to walk through real examples, answer questions, and let employees actually try things out. That part really mattered. One of the biggest “misses” I see with AI rollouts is assuming everyone already knows how to use it or, on the flip side, assuming people will just “figure it out.” (They won’t. And also… please don’t upload company financials.)

Before you even get to a launch, I strongly recommend starting with an AI policy. This doesn’t have to be scary or overly legal, but it does need to be clear. Outline the dos and don’ts, what’s okay to input, what’s off-limits, and any department- or use-specific guidance. The goal is to remove ambiguity so people feel confident using AI instead of nervous or reckless.

From there, short trainings go a long way. Think quick sessions, practical demos, and real use cases that show how AI can actually help people in their day-to-day work. Follow those up with regular touchpoints or reminders so the guidance stays fresh and top of mind — especially as tools and use cases evolve.

Responsible AI rollout is really about sequencing — clarity first, then speed.

This is also where internal comms plays a really important role. When leadership is pushing for speed, comms can help set the pace. That means reinforcing expectations, repeating the guardrails, and creating space for questions and feedback as people start using these tools in real life.

Honestly, introducing AI responsibly is less about getting it perfect on day one and more about being clear, being available, and being willing to adjust as you go. If people know what’s okay, what’s not, and who to ask when they’re unsure, you’re in a pretty good place. The rest tends to sort itself out.

Devin Owens signature
Devin Owens

Hey there, I'm Devin!

Most of the time you can catch me deep in the world of internal comms at Workshop (yes, the Happy Mondays folks!), and while I love AI, there are just some comms questions that need a human who really gets it… that’s me!