Most destinations have either handed their judgment to AI or banned it outright. The answer sits in between.

When wildfires forced closures at the Grand Canyon, the fire wasn’t the only problem. Even after parts of the park reopened, visitation stayed low because people assumed the whole park was still closed.

So we added AI-generated “Now Open” signs to real photography of the canyon. Nobody believed those signs were sitting on the rim, and nobody was supposed to. The real landscape still did the selling.

That campaign is a good picture of where AI belongs in destination marketing. Right now, many destination marketing organizations (DMOs) aren’t there yet. They’re still figuring out where AI fits in their work, and many have landed at one of two extremes.

The first is leaning too heavily on the tools. Teams accept AI outputs at face value in place of their own strategic reasoning, which Steven Shaw and Gideon Nave at the Wharton School call cognitive surrender. When that happens, the quality of the work slips, and the team loses control over what it puts in front of travelers.

The second is an outright ban. Some organizations overcorrect with a simple policy: no AI. That protects them from one set of mistakes, but it also shuts them out of what AI does well.

I’m not saying you should use AI for everything. I’m saying its real value sits between those two extremes. Used well, AI informs your judgment. It surfaces opportunities you hadn’t recognized and shows you gaps in your thinking. It helps you make smarter decisions, like where to redirect visitors when a destination gets overcrowded or how to respond when world events start shifting travel patterns.

The most effective AI policies are built around that judgment. They’re specific about AI’s limits and open to the analysis it does so well.

Use AI to find the patterns you can’t see

AI is excellent at recognizing patterns and finding correlations between pieces of information that seem unrelated. That makes it useful anywhere a problem is logical, from building a pricing structure to pressure-testing the assumptions in your plan.

For a DMO, one of the most valuable applications is seeing what’s coming. Much of what shapes your visitation sits outside your control and outside your marketing data. Let’s say a competing destination starts courting your core visitors with an offer that feels newer or better timed than yours. Or geopolitics shifts how international travelers feel about a trip to the United States. Or tariffs raise the cost of the gear your visitors need, and fuel prices change the math on a road trip.

Most of those forces leave a traveler’s motivation intact. In Behavioral Marketing terms, what they add is friction to the traveler’s ability to act on it. Often, the first sign is a dip in your booking data, and it arrives without an explanation. You may not have known anything was shifting, let alone what was driving it.

Each of those signals lives in a different place, and connecting them is the kind of correlation work AI does well. It can show you that travel from a particular country or audience is likely to drop while there’s still time to adjust your audience mix or your messaging. We’ve made the case for treating policy signals as marketing intelligence, and AI can do a lot of that watching for you.

Policy signals are just one example. AI can sharpen any decision that depends on spotting patterns in more information than your team can work through on its own.

Keep human judgment between the tool and the traveler

That said, analysis is only half the picture. The other half is what you put in front of travelers, and that’s where cognitive surrender does the most damage.

The fix is to make sure a person’s judgment sits between the tool and any finished work. Every time. No exceptions.

In destination marketing, that standard matters most for audience-facing materials, especially imagery and copy.

Use AI in your imagery transparently or not at all

Keep AI imagery to a few clearly defined uses. Whatever you do, don’t use AI to fabricate images of your destination, and don’t enhance a real image past what a visitor is likely to see. Both are false, and both set up expectations your destination can’t meet.

Travelers measure the place in front of them against the pictures they saw before they booked. When those pictures were generated or overworked, the gap shows up the moment they arrive. The trust you lose is hard to win back.

AI still has a place in your visual work, as long as nobody could mistake the result for the real thing. That’s why the “Now Open” signs worked. They were clearly added to the scene, and everything behind them was real.

Use AI to strengthen your copy, and rely on people for the experience

Copy is a less obvious risk. AI-drafted content can look polished and still miss what makes a place worth visiting. You can load your brand voice into an AI tool. That doesn’t mean it will get the voice right.

I’ve hiked the Grand Canyon rim to rim nine times. I’ve spent seven days rafting the river through it, swimming in that river, running its trails, and riding four hours by four-wheel drive to reach Cape Solitude. When I write about the smell of the place or what it feels like to stand at the edge, I’m drawing on all of that. AI is working from what other people have written about it.

That still leaves plenty of room for AI in the drafting process. A tool can produce a solid first draft quickly, and handing it that kind of execution is efficient. What it can’t do is decide whether the draft is right. Treat the draft as raw material, and review every line with the same care you’d give a new writer’s work. Use the tool to suggest angles you haven’t considered or to pressure-test what you’ve written. Then let a writer who knows the place bring in the details only experience can supply and make the final call on what goes out.

Your approach to AI won’t look like anyone else’s

The right balance of AI always puts human judgment first. That’s as true for DMOs as it is for any other brand, but what it looks like in practice will be different for every organization.

What won’t work is either extreme. Cognitive surrender removes the judgment that makes your work worth trusting. Avoiding AI altogether means giving up the patterns and early warnings it can surface. Hold on to either one, and you risk falling behind the destinations that use these tools wisely.

So whatever your current approach, ask the same question: Where does a person’s judgment need to sit between the tool and the finished work? Answer it for each part of your work, from research to copy, and you’ll know where to lean in and where to hold the line.

Remember those “Now Open” signs. AI put them on the photos, but a person decided they belonged there. I’ve spent enough time in that canyon to know no generated image could replace it. AI didn’t need to. It only needed to tell people the real one was open again.

Off Madison Ave builds Behavior Blueprints for destinations and hospitality brands. To see why AI-generated imagery puts traveler trust at risk, read AI Can Make Your Destination Look Stunning. And, That’s The Problem.