We're All Using AI. Let's Talk About it.

I facilitated a conversation about AI in strategic communications recently and I'll tell you exactly how I opened the session. "I'm not an expert. I want to be upfront about that."

A room full of senior tech communications and PR professionals — most of them with 20-plus years in the industry — and that was my opener. Not a confidence move in the traditional sense. But the right move.

Here's why.

The expert trap

There's a version of AI training that positions the facilitator as the all-knowing authority, there to download information to a room full of people who don't know any better. That model doesn't work in a room like this one — and frankly, it doesn't work in most rooms anymore.

The people I was with this week have spent decades developing instincts that AI can't replicate: how to read a room, how to counsel a CEO through a crisis, how to land a complicated story in a noisy media environment. What they didn't always have was fluency in the tools that are changing how all of that work gets done.

My job wasn't to be the expert. My job was to be the translator and to make space for everyone to bring what they already knew to the conversation.

Where we started

We built from the foundation up and I mean the actual foundation, not the surface-level stuff most AI trainings skip.

We covered how LLMs actually work: the stack from models to tools to platforms to agents, and why understanding that distinction changes how you deploy them. We talked through why paid beats free — and why Pro ≠ Enterprise, which matters enormously for a team operating inside a company with strict data governance and the agencies that support them. We walked through the only security question that actually matters: Is this already public? That single filter cuts through more anxiety and confusion than any policy document.

We got into which tools are built for what. Claude for long-form synthesis and structured output. ChatGPT for speed and versatility. Grok, uniquely, for real-time X/Twitter intelligence. Perplexity for live, cited research. Not trivia. Operational fluency.

And we talked about shadow AI and BYOAI — because eight out of ten people in that room were already using tools their IT team didn't know about. That's not recklessness. That's what happens when the official path is worse than the unofficial one.

Then we went live

The foundation matters, but the real unlock is in application.

We took scenarios these communicators face every single day and worked through them in real time. With the help of a quick advance survey, we talked through the things the team admitted spending hours on. Live prompts. Live refinements. Live outcomes.

That part always changes depending on who's in the room. The tools and the frameworks stay consistent. The use cases get built around what your team actually does — your media mix, your workflows, your pressure points.

And every time, there's a moment where the room shifts. Not because it's magic. Because it's suddenly theirs.

What I actually believe about AI in communications

You can't iterate at the speed of AI by going it alone, and you can't pretend you've figured it all out. The field is moving too fast. The tools are changing too quickly. Anyone who tells you otherwise is either selling something or not paying close enough attention.

What you can do is come together with people who are using these tools in the same contexts you are, compare notes honestly, and build collective fluency. That's how AI gets out of the shadows. That's how teams actually level up. Not by mandating tools from the top or waiting for an official training program, but by creating space for practitioners to talk to each other about what's working.

Every room teaches me something. This one was no different — and it was one of the best conversations I've had about where we are and where we're going.

Let's do more of these

If your communications team is ready to move from AI-curious to AI-proficient, I build and facilitate AI communications workshops and playbooks designed specifically for comms teams — grounded in the work you're actually doing, not generic AI training. We cover the tools, the security questions, the prompting frameworks, and the live use cases that change how your team operates.

Get in touch → carey@careyworks.com