The precision gap: Why human expertise still trumps AI in international travel

When Rita Santos asked a popular AI chatbot if she could get a visa at the Vietnam/Cambodia land border, the response was instant and confident. It was also completely false.

This “hallucination”, a common flaw where Large Language Models (LLMs) invent facts, didn’t just ruin Rita’s itinerary; it left her stranded in 40°C heat. While the AI’s outdated data caused the crisis, it was only through the intervention of a professional visa agency that she was able to secure emergency documentation and continue her journey.

This story highlights a growing danger in the travel industry: the gap between AI-generated convenience and vetted, professional accuracy.

As general-purpose LLMs become a default resource for travel planning, iVisa, a global leader in travel documentation, is issuing a formal warning: Unverified AI chatbots are structurally unsuited to provide regulated legal requirements. Without real-time verification and human oversight, the consequences of trusting automated ‘guesses’ can be severe.

When “fast” information fails

Rita Santos is a solo traveler currently on a mission to visit every country in the world. At a land border crossing between Vietnam and Cambodia, she asked a popular AI chatbot to confirm whether a visa could be obtained on arrival at that specific location. The chatbot confirmed it could.

It could not.

Upon arrival, Rita discovered the border did not issue visas on arrival for her nationality. The cascade of consequences was immediate.

• Financial: A forced $250 emergency fee to a third-party agency for a rushed visa
• Physical: Three hours waiting in a 40°C heat, leading to near-fainting and physical exhaustion
• Practical: Stranded alone at a remote crossing with no support desk, no recourse, and no way to recover the costs incurred

“Wrong information at the wrong moment has a real cost,” Rita says. “I wasn’t just dealing with a logistical nightmare, I was in survival mode. You need a tool you can actually trust when you are at your most vulnerable.”

Rita’s account is one incident. But it points to a pattern that travel professionals and immigration specialists are increasingly flagging as both common and preventable. According to a February 2025 Kantar survey of more than 10,000 consumers across ten countries, 40% of global travelers have already used AI tools to plan their trips, with 62% open to doing so in the future.

Why AI falls short for travel logistics

iVisa identifies three primary reasons why LLMs are currently unsuitable for visa research:

1. Stale data: AI models are trained on historical datasets with a fixed knowledge cutoff. Visa policies can change overnight due to diplomatic shifts, new government mandates, or public health directives. With no guarantee that any given AI model reflects current requirements.
2. Lack of nuance: Entry requirements frequently differ depending on the specific port of entry. Visa-on-arrival policies at international airports often do not apply at land border crossings. AI models routinely conflate the two, with no mechanism for flagging the distinction.
3. Zero accountability: If an AI provides incorrect visa advice, there is no support desk to contact and no way to recoup financial losses. The AI industry currently has no obligation to disclose the limits of its immigration knowledge or direct users to authoritative sources.

The solution: Human-verified accuracy

Unlike general-purpose AI tools, specialist visa platforms maintain connections to government immigration databases, employ dedicated regulatory monitoring teams, and update entry requirements in real time. The difference between a knowledge cutoff and a live data feed is, in practice, the difference between last year’s visa policy and today’s.

“Travel is about freedom, but that freedom rests on the foundation of accurate legal documentation,” says Victor Gimenez, Director of Customer Experience at iVisa. “We don’t guess; we verify. Our platform combines cutting-edge technology with a global team of experts who monitor government updates in real-time.”

Advice to travelers

iVisa urges all international travelers to take the following steps before any border crossing:

• Treat AI-generated visa information as a starting point, never a final answer: Cross-reference any chatbot output against official government sources or a verified specialist platform.
• Check requirements specific to your nationality, destination, and entry point; Rules vary significantly between airports, land borders, and ports. And between passport types.
• Verify close to your departure date. Regulations can change with limited notice. Information accurate at the time of booking may not be accurate when you travel.
• Use sources that update in real time. Government immigration portals and specialist visa services maintain current data. General-purpose AI tools do not.