Is ChatGPT Putting Your Traveling Employees at Risk?

Travel risk managers and corporate security teams need to understand the risks involved with employees relying on ChatGPT to plan travel.

On LinkedIn, Stuart McDonald, the founder of Travelfish.com, shared an article from the Verge entitled, “A Bride, a Groom, a Honeymoon, and ChatGPT.

The article begins with “As the internet took over, millennials ditched travel agents in favor of a DIY approach. With AI, they’re rediscovering the joys of letting someone — or something — else do the planning.”, and discusses how people are turning to ChatGPT to design travel itineraries.

In my business school and college teaching, I’ve experienced many of ChatGPT’s flaws. Students have even failed classes thanks to ChatGPT. They didn’t fail simply because they used the Openai site, they failed because of all the mistakes the language bot made that the students didn’t notice and correct before submitting assignments. E.g., one assignment looked to be very well written, and it contained numerous citations of scientific works. The only problem was that the citations were made up. The quotes were not found in any of the cited articles…

In one of my classes today, I gave my students a simple exercise. Use ChatGPT to decide if Grindavik, Iceland is a nice place to visit on holiday.

The bot told them that, “Grindavik, Iceland, is indeed a fantastic destination with a unique landscape and various activities to enjoy.” It also recommended ten fun activities for tourists before concluding, “Remember that Iceland’s weather can be unpredictable, so it’s essential to be prepared for various conditions when exploring the outdoors. Whether you’re interested in relaxation, adventure, or cultural experiences, Grindavik and the surrounding areas offer a diverse range of activities for a memorable vacation.”

In other words, to safely enjoy Grindavik, bring an umbrella.

The students all agreed it would be cool to visit Grindavik before they did part two of the assignment which was to search online media sites for news about Grindavik.

After learning that the village had been evacuated and homes destroyed by volcanic eruptions and lava flows, they understood that taking an umbrella along on their holiday would not help them enjoy the ghost town with restricted access Grindavik had become.

What does this mean for the roles of travel managers, travel risk managers, and security teams that oversee corporate travel?

Firstly, they need to be aware that many of the people who travel on behalf of their companies may be putting themselves at unnecessary risk by using ChatGPT with its outdated information. Even worse, perhaps some of them are guilty of using it to produce destination information that is provided to travelers.

Travel risk is more than keeping up with the news.

Early in my corporate career, before we focused on travel risk management, I was sent to visit a hotel General Manager in a very remote location.

I asked him what he knew about the country before he arrived.

“Nothing”, he said. “My boss phoned me and told me if I wanted to be a GM I could start here. When I told him I didn’t know anything about the place, he said not to worry. When I picked up my tickets at the airport, they gave me a guidebook. After takeoff, I opened the book and the first thing I read was that this is a Muslim country. I’m Jewish!”

Fortunately, the GM was treated very well in his new country, but he had been shaken when he read those lines and even more so when, after being picked up at the airport, the driver stopped outside a synagogue and said, “That’s where you can go to pray.” The driver knew more about him than he knew about the country he would be living in.

Travel security is personal.

Regarding travel, algorithms and bots may be pretty good at generically telling us things we want to see, but they’re not great at tailoring guidance to protect the specific risks a company or an individual may face.

To Stuart McDonald’s post on LinkedIn, I offered the following comment:

“It’s simple really, ask a bot that only knows algorithms, or ask a person who knows both you and where you want to go on holiday.”

At Risk Resiliency, we have decades of experience providing guidance and support to people travelling to and living in hundreds of countries. We have built trusted networks of people who can provide up-to-date, first-hand knowledge of issues companies or their travelling personnel need to know.

Who would you place your trust in?

A bot full of outdated information and algorithms or a person with experience, expertise, and an abundance of personal contacts around the world?

The choice is yours!


Questions for Consideration

  1. Are your warning your corporate travelers against relying on ChatGPT for travel advice?
  2. Do you share destination cautions and safety advice with your travelers?

Need help with your travel security program? We’d love to chat (the real kind, not the GPT kind). 


Recommended Reading

Protecting Business Travelers

People are any company’s most valuable asset and as such protecting them at work should be a prime objective; this is perhaps more so when companies send their people travelling . . .

Read More

The Ten Pillars of Hospitality Security

Organizing your resilience efforts around ten pillars and then assessing your efforts systematically will make sure your programs are achieving world-class results.  Pillar 1: Threat and Intelligence Pillar 2: Program . . .

Read More