I need to talk to you about something I’m seeing way too often in cruise Facebook groups, travel forums, and even in conversations with my own clients.

People are trusting AI chatbots like ChatGPT, Grok, Gemini, and Claude to plan their vacations. And it’s leading to disaster.

Look, I get it. These tools are impressive. Ask ChatGPT to write you a poem and it’ll blow your mind. Ask it to explain quantum physics to a five-year-old and it’s genuinely helpful. But ask it about your Virgin Voyages cruise? About flight prices? About visa requirements?

That’s where things go terribly, terribly wrong.

The Virgin Voyages Misinformation I’m Seeing Every Day

Let me share some real examples of what AI chatbots are telling sailors about Virgin Voyages—and why it’s flat-out wrong.

❌ “Your Deck 11 cabin has an obstructed view from lifeboats”

I’ve had clients come to me panicked because ChatGPT or another AI told them their Sea Terrace cabin on Deck 11 has lifeboats blocking the view. This is completely false. Virgin Voyages’ ship design places lifeboats on lower decks. The AI is apparently confusing Virgin Voyages with other cruise lines or just making things up entirely.

❌ “You’ll need to pay extra for The Wake steakhouse and other specialty dining”

Wrong again. One of the most incredible things about Virgin Voyages is that all 20+ eateries are included in your fare—including The Wake, Razzle Dazzle, Pink Agave, Gunbae, and every other restaurant onboard. There are no specialty dining upcharges. The AI is applying outdated cruise industry norms to a brand that threw out the rulebook.

❌ Cruise prices that are thousands of dollars off

This one genuinely scares me. People are being quoted prices by AI chatbots that are sometimes $2,000 or more below actual fares. They budget based on these fictional numbers, get excited, and then experience crushing disappointment when they see real pricing. AI tools don’t have access to live cruise pricing—they’re essentially guessing based on outdated training data.

We Tested ChatGPT for Flight Booking. It Was a Disaster.

I keep seeing posts in travel groups telling people to “STOP buying plane tickets from airlines and go to ChatGPT to find the best flights and prices!”

So I decided to test it myself. Here’s what happened:

• Flight days were wrong — It suggested flights on days they don’t operate
• Itineraries were impossible — Connections that don’t exist, layovers in the wrong direction
• Prices were completely fabricated — Numbers that had no basis in reality

One travel publication ran a similar test and found ChatGPT offered a 29-hour flight via Atlanta and Detroit for a route that has multiple direct flight options taking a third of the time. It insisted there were no direct flights available. There were plenty.

Even worse? Google’s Bard (now Gemini) once suggested a hotel in Tokyo called the “Hotel Gracery Shibuya” for a user looking to stay in Shibuya. The hotel doesn’t exist. It was completely fabricated—an AI “hallucination” that would have left a traveler stranded.

Real People, Real Consequences

This isn’t theoretical. These AI mistakes are ruining vacations and costing people real money. Here are documented cases:

The Spanish Couple Denied Boarding Because of ChatGPT

In August 2025, Spanish travel influencer Mery Caldass and her boyfriend were traveling to Puerto Rico when they were stopped at the airport and denied boarding. Why? They’d asked ChatGPT if they needed a visa, and it said no. What ChatGPT failed to mention? Spanish citizens need to apply for an Electronic Travel Authorization (ESTA) before traveling to Puerto Rico.

The tearful video of Caldass being consoled at the airport went viral on TikTok, with millions of views. In the video, she says through tears: “I always do a lot of research, but I asked ChatGPT and it said no.”

AI doesn’t understand the nuances of travel requirements—and the difference between “no visa” and “ESTA required” cost them their entire trip.

The Australian Author Stranded in Mexico City

In March 2025, Australian author and strategist Mark Pollard was on a lecture tour across Latin America. He’d used ChatGPT successfully for months and had come to trust it. When he asked which countries in Latin America required a visa for Australians, it told him Chile didn’t require one—Australians could stay up to 90 days visa-free.

Wrong.

At the check-in counter in Mexico City, Pollard learned the devastating truth: Australian citizens have needed a visa for Chile since 2019, and it requires a 20-day processing period through the Chilean consulate. His conference was cancelled. His plans were destroyed. He asked ChatGPT if he could sue it. (It said no.)

His post about the incident generated over 15 million views on Instagram. As he later explained, he’d developed “excessive confidence” in AI after months of helpful interactions—exactly the trap many travelers fall into.

Air Canada Loses Tribunal Case Over Chatbot Lies

When Jake Moffatt’s grandmother passed away on November 11, 2022, he immediately visited Air Canada’s website to book a flight from Vancouver to Toronto for the funeral. While researching bereavement fares, he used Air Canada’s support chatbot, which told him he could book a regular ticket and submit for a reduced bereavement rate within 90 days.

This was completely false. Air Canada’s actual policy doesn’t allow retroactive bereavement fare requests.

Moffatt booked flights totaling over $1,600 CAD, submitted his claim with his grandmother’s death certificate, and was denied. When he sued, Air Canada tried an extraordinary defense: they argued their chatbot was a “separate legal entity that is responsible for its own actions.”

The tribunal member called this “a remarkable submission” and ruled against Air Canada, noting: “It should be obvious to Air Canada that it is responsible for all the information on its website. It makes no difference whether the information comes from a static page or a chatbot.”

Air Canada was ordered to pay $812 CAD in damages and fees. The case made international headlines as a warning about AI misinformation in customer service.

When AI Gets It Dangerously Wrong

These aren’t just inconveniences. AI hallucinations can be genuinely dangerous:

Tourists in Peru arrived in a rural town determined to hike to a “sacred canyon” their AI chatbot recommended. The canyon doesn’t exist. A high-altitude hike in an area with spotty cell coverage could have been life-threatening if they’d attempted it.

A couple in Malaysia made the trek to see a scenic cable car they’d seen in what appeared to be a travel video. When they arrived, they discovered the cable car—and the entire destination—was AI-generated. It never existed.

In Japan, one AI tool told travelers that the last ropeway down from a mountain was at 5:30 PM. The actual last departure was much earlier. They found themselves stranded at the mountain top as darkness fell.

The $100 Billion Mistake

Even the companies building these AI tools can’t control their accuracy.

When Google unveiled its AI chatbot Bard (now Gemini) in February 2023, they showed it answering a simple question: “What new discoveries from the James Webb Space Telescope can I tell my 9-year-old about?”

Bard confidently responded that the telescope “took the very first pictures of a planet outside of our own solar system.”

Completely wrong. The first exoplanet images were captured by the European Southern Observatory’s Very Large Telescope in 2004—nearly two decades earlier.

This single error, in Google’s very first public demo, caused Alphabet’s stock to plummet. The company lost over $100 billion in market value in a single day.

If Google can’t even fact-check their own demo, why would you trust AI to plan your vacation?

Lawyers Are Getting Sanctioned. Travelers Could Be Next.

Here’s something that should terrify anyone relying on AI for important decisions:

A New York lawyer named Steven Schwartz used ChatGPT to research legal cases for a brief he was filing in federal court. ChatGPT provided him with six cases that supported his argument perfectly—complete with case numbers, quotes, and citations.

None of them existed.

ChatGPT had invented entirely fictional court cases, complete with fabricated judges, made-up quotes, and fake case numbers. The lawyer was fined $5,000 and publicly humiliated. The judge noted that ChatGPT had created “bogus judicial decisions with bogus quotes and bogus internal citations.”

Since then, lawyers across the country—and around the world—have faced sanctions for citing AI-fabricated cases. One research tracker found that cases of AI hallucinations in legal filings went from about two per week in early 2025 to two to three per day by fall.

If lawyers—trained professionals with access to sophisticated research tools—are falling for AI hallucinations, what chance does the average traveler have?

Why AI Gets Travel So Wrong

Understanding why this happens might help you avoid the trap:

AI doesn’t “know” anything. Large language models like ChatGPT, Gemini, and Claude don’t have databases of facts they look up. They predict the most statistically likely next word based on patterns in their training data. They’re essentially very sophisticated autocomplete.

Training data is outdated. Most AI models are trained on data that’s months or years old. Cruise policies change. Airlines update routes. Visa requirements shift. The AI has no idea what’s current.

AI can’t access live data. Unless specifically connected to booking systems (which most general-purpose chatbots aren’t), AI cannot see real prices, availability, or current policies. Every number it gives you is essentially a guess.

AI is designed to always give an answer. These tools never say “I don’t know.” They’re programmed to provide confident, helpful-sounding responses—even when they’re completely making things up. As one AI expert put it, they’re “little puppies who want to please you, and will hallucinate and lie to give you some kind of answer.”

What AI CAN Help With (And What It Can’t)

I’m not saying AI is useless for travel. Here’s a realistic breakdown:

Will AI Get Better? Probably. Is It Ready Now? Absolutely Not.

Will AI eventually be able to reliably help with travel planning and booking? Almost certainly. Companies are working on connecting AI to live booking systems, verifying information in real-time, and reducing hallucinations.

But as of December 2025? We’re not there yet.

Even the best AI models still hallucinate at rates that would be completely unacceptable for travel planning. One recent study found that even top-performing models make things up in about 3% of responses—and that rate jumps dramatically for specialized topics like legal information (6.4%) or fast-changing data like travel policies.

A 3% error rate might sound small. But if you’re planning a two-week vacation with dozens of details—flights, hotels, transfers, activities, dining reservations, visa requirements—even a small error rate means something is probably wrong with your trip.

The Human Amplification Problem

Here’s something that makes this even worse: the misinformation problem isn’t just AI anymore.

Someone asks ChatGPT a question about Virgin Voyages, gets a wrong answer, and then shares that “fact” in a Facebook group or Reddit thread. Other sailors read it, believe it, and share it further. Before you know it, a complete AI hallucination has become “common knowledge” in the cruise community.

But AI isn’t the only source of bad information. I wrote an entire article about the dangers of online forums: The Siren’s Call: Why Facebook & Reddit Can Sink Your Virgin Voyages Dream. In it, I describe online forums as “the Wild West of travel planning”—filled with well-intentioned people, but also rife with misinformation that can sabotage your experience.

The sources of bad information are everywhere:

• AI-Sourced “Expertise”: That confident, well-written response about Virgin Voyages policies? It might have been copy-pasted directly from ChatGPT by someone who didn’t bother to verify it.
• Angry Travelers with an Axe to Grind: Someone had one bad experience—maybe their steak was overcooked or they didn’t get the cabin upgrade they wanted—and now they’re on a mission to trash the entire brand. They exaggerate, embellish, or outright fabricate “facts” because they’re upset and want others to share their disappointment.
• ️ Competitors in Disguise: This one’s insidious. Other cruise lines, competing travel agents, or people with financial interests will pose as regular travelers and spread false information to steer you away from Virgin Voyages. That “helpful” person warning you about made-up problems? They might be selling something else entirely.
• The Echo of Outdated Information: A post from 2022 about a specific promotion gets shared as gospel in 2025. Now imagine that “advice” originally came from ChatGPT, which was trained on even older data.
• The Reddit Rabbit Hole: You’ll read contradictory “facts” about cabin upgrades, dining reservations, and policies—some from actual sailors, some from angry people making things up, some regurgitated from AI tools, and no way to tell the difference.

Virgin Voyages is a dynamic brand that’s constantly evolving. Bar Tab bonuses change. New dining concepts are added. The VoyageFair Choices fare system just launched in October 2025. Loyalty perks get updated. Advice that was accurate six months ago might be completely wrong today—and whether the source is AI hallucinations, bitter travelers, or hidden competitors, the result is the same: you end up with bad information that can wreck your vacation.

The Human Difference

So, if AI chatbots are a minefield and online forums are the Wild West, where’s the lighthouse? It’s your dedicated, professionally trained travel advisor.

Here’s what a human travel advisor (like me) can do that AI cannot:

• ✨ Access live booking systems with real-time pricing and availability
• ✨ Verify information against official sources before giving you advice
• ✨ Understand context — your specific needs, preferences, and concerns
• ✨ Take responsibility — if something goes wrong, there’s a real person accountable
• ✨ Stay current — I know about the VoyageFair Choices changes, the new dining reservation windows, and every policy update
• ✨ Know what they don’t know — and tell you honestly when they need to check on something

Most importantly? My services don’t cost you anything extra. As a Virgin Voyages travel advisor, I’m paid by the cruise line—not by you. You get expert guidance, personalized service, and accurate information at no additional cost.

When things go wrong, a Facebook group can offer sympathy. ChatGPT can offer a confident-sounding hallucination. A First Mate with integrity offers solutions.

The Bottom Line

AI tools are incredible for many things. Planning your vacation is not one of them—at least not yet.

Every week, I help clients who came to me after AI chatbots gave them bad information. Clients who budgeted based on fictional prices. Clients who worried about non-existent problems. Clients who almost missed out on the trip of a lifetime because a robot told them something that wasn’t true.

Don’t let that be you.

If you’re planning a Virgin Voyages cruise—or any cruise—work with someone who has access to real information, real booking systems, and real accountability. Someone who’s sailed Virgin Voyages multiple times, achieved Gold Tier First Mate status, and been recognized as one of the Top 100 First Mates in North America.

Someone who won’t hallucinate your vacation into disaster.

Frequently Asked Questions

Can I use ChatGPT to help plan my cruise at all?

You can use it for very general brainstorming—like “what are some things couples enjoy doing in the Caribbean?” But for anything specific about Virgin Voyages policies, pricing, cabin details, or itineraries, always verify with official sources or a qualified travel advisor. ChatGPT does not have access to live cruise pricing or current policies.

Why does AI make up information instead of saying “I don’t know”?

AI language models like ChatGPT, Gemini, and Grok are designed to be helpful and always provide an answer. They predict likely responses based on patterns in their training data, not verified facts. When they don’t have accurate information, they generate plausible-sounding text instead of admitting uncertainty. This is called “hallucination” and is a fundamental limitation of current AI technology.

Is ChatGPT good for finding cheap flights?

No. ChatGPT cannot access live flight pricing or availability. When tested, it provided flights on wrong days, impossible itineraries with connections that don’t exist, and completely fabricated prices. One test found ChatGPT suggesting a 29-hour flight for a route with multiple direct options. Always use actual flight booking sites like Google Flights, airline websites, or a travel advisor.

Can I trust ChatGPT for visa requirements?

Absolutely not. There are documented cases of travelers being denied boarding because ChatGPT gave them incorrect visa information. In 2025, Spanish tourists missed their flight to Puerto Rico because ChatGPT didn’t mention ESTA requirements, and an Australian was stranded in Mexico City because ChatGPT was wrong about Chile’s visa policy. Always check official government sources for visa requirements.

What is an AI hallucination in travel planning?

An AI hallucination is when an AI chatbot generates false information that sounds confident and plausible but has no basis in reality. In travel, this includes fake hotel names, non-existent destinations, incorrect visa requirements, fabricated prices, and wrong flight schedules. Google’s Bard once suggested a hotel in Tokyo that didn’t exist, and travelers have been directed to “sacred canyons” and cable cars that were completely made up.

Does Virgin Voyages charge extra for specialty dining?

No—and this is a common AI mistake. AI chatbots often claim Virgin Voyages charges extra for specialty restaurants like The Wake steakhouse. This is completely false. All 20+ eateries on Virgin Voyages ships are included in your fare—The Wake, Razzle Dazzle, Pink Agave, Gunbae, and every other restaurant. There are no specialty dining upcharges. AI often applies outdated cruise industry norms to Virgin Voyages incorrectly.

Are any AI travel tools reliable for booking flights or cruises?

Some travel companies are building AI tools connected to their own verified databases, which can be more reliable for that specific company’s information. However, general-purpose chatbots like ChatGPT, Gemini, Grok, and Claude should never be trusted for specific travel facts like visa requirements, current prices, flight schedules, or cruise policies. Always verify with official sources.

What should I do if AI gave me travel information I already acted on?

Immediately verify that information with official sources—airline websites, cruise line websites, government visa pages, or a qualified travel professional. If you’ve already booked something based on bad AI information, contact the provider directly to understand your options. In some cases, like the Air Canada chatbot lawsuit, companies have been held liable for AI misinformation.

Why should I use a travel advisor instead of AI for cruise planning?

Human travel advisors have access to live booking systems with real-time pricing and availability, can verify information against official sources, understand your specific needs and context, take responsibility if something goes wrong, stay current on policy changes, and know when they need to check on something. Services from Virgin Voyages specialists like CamJon Travel are free—travel advisors are paid by the cruise line, not by you.

Will AI travel tools get better in the future?

Almost certainly. Companies are actively working on reducing hallucinations and connecting AI to live data sources. But as of December 2025, no general-purpose AI chatbot should be trusted for specific travel planning decisions. The technology is improving but not yet reliable enough for high-stakes travel bookings.