When any substantial change captures our attention—whether it’s a trend in fashion, a new business model, or a cultural movement—a new language often emerges. From that language comes a vocabulary, and that vocabulary eventually spreads to the point that many of us adopt it.

But sometimes words spread faster than meaning. We find ourselves using them without really knowing what they mean.

Nowhere is this more apparent right now than in the vocabulary of AI.

AI-First

Take the phrase “AI-first.” If you’re anything like me, you’ve read and heard it countless times over the last year. And if you’re anything like me, your understanding of it has been, at best, a little vague.

That lack of clarity says less about us than about the speed of this moment. The truth is, there isn’t a single agreed-upon definition. Which means, inevitably, it’s up to each of us—and each organization—to decide what AI-first really means.

Back in 2021, before ChatGPT, startup investor Ash Fontana published The AI-First Company. He wrote: “AI-first companies put AI to work, prioritizing it within real budgets and time constraints. AI-first companies make short-term trade-offs to build intelligence in order to gain a long-term advantage over their competitors.”

Useful, but incomplete. The costs of building intelligence from the data you already own are falling so rapidly that the “short-term trade-offs” he describes are no longer the obstacle they once were. And so, even with a book-length treatment, we’re left without a definitive meaning.

Which brings me back to the present moment, and to my own definition.

AI-First or Strategy-First?

In the brand and marketing work I do, I’ve long insisted on strategy-first. Strategy before execution, always. Clients sometimes resist this—speed feels easier than precision—but the principle is simple: if the strategy is wrong, execution won’t matter.

That doesn’t change in the AI Age. If anything, it becomes more important. Because what we’re really talking about here are mindsets. Strategy-first is a mindset. AI-first is, too.

To declare yourself AI-first is to signal intent—to your team, your market, and yourself. It says you understand the sea change happening around us, and you’re not watching it from the shoreline. You’re stepping into the water.

What you’re not saying is that AI will do everything. AI-first isn’t absolutist. It’s not a destination. It’s a direction.

Always Invite AI to the Table

Wharton professor Ethan Mollick, in Co-Intelligence, urges leaders to “always invite AI to the table.” His point isn’t that AI always helps. Sometimes it frustrates, sometimes it misfires. The point is that by inviting it in, you learn where it fits—and where it doesn’t.

This echoes AI expert Lance Eliot, who argues: the more you can think like the machine, the better you can contend with it. To be AI-first, then, is to be fluent enough to work with the machine without forgetting you’re human.

My Own AI-First

For me, AI-first doesn’t mean I reach for AI before anything else. It means I reach for it as part of the process. I use it to sharpen strategy, to pressure-test ideas, to draft, analyze, and brainstorm. Sometimes the output is useful, sometimes it isn’t—but the value lies in the act of including it, in making AI a regular companion to the work rather than an afterthought.

Which brings me back to the vocabulary itself. New words often resist strict definition. Sometimes it’s easier to understand their intent than to pin them down precisely.

AI-first is one of those words. It isn’t a rigid formula. It’s an orientation. A choice. A mindset that says: AI belongs in the process—not replacing human judgment, but working alongside it.

And that may be all the definition we need.