‘Don’t make my mistake’: Aussie traveller’s border nightmare after relying on ChatGPT

With artificial intelligence apps fast becoming more common, it can be tempting to use them to get travel advice.

However, many holidaymakers are finding out the hard way they’re not always accurate.

When Max Phelps, 58, from Sydney was planning a once-in-a-lifetime trip to Africa, he logged onto popular AI program, ChatGPT.

But the advice it gave left the money coach and mortgage broker, who was holidaying with girlfriend, Julie and grandson, Matthew, scrambling.

“My girlfriend and I knew that we wanted to see Victoria Falls and turned to AI for the practicalities, like the best places to see it from, where to stay, and where to fly into,” he told nine.com.au.

Read more

‘Yes, there is snow!’ Aussie ski season officially in full swing

“AI recommended seeing Victoria Falls from both the Zimbabwean and Zambian sides, then a safari in Botswana’s Chobe National Park.

“So we built our itinerary around that.”

ChatGPT told him to get the KAZA Univisa, which it claimed covered travel across all three countries with unlimited re-entry for 30 days.

He made the mistake of not cross-checking this information.

As Aussies, they didn’t need a visa for Botswana or Zambia, only Zimbabwe. And the visa they got wasn’t correct.

“The problem with the KAZA Univisa is that while it allows unlimited travel between Zimbabwe and Zambia, it becomes invalid if you leave those two countries for one night or more,” he said.

“It wasn’t until we tried to re-enter Zimbabwe after the safari that border officials told us our visa was no longer valid.”

Luckily, they were able to buy a visa at the border.

But Max said it took time, caused stress and cost $43 each in visa fees, on top of the $72 each they’d paid.

Read more

‘Behind a hill was the most incredible hotel room I’ve ever stayed in’

He warned other travellers not to make the same mistake.

“If you’re using AI to plan a trip, don’t just take its first answer as the complete picture, especially for anything involving visas, borders, or legal requirements,” he said.

Research from Send My Bag found nearly four in 10 Australians have used AI to research or plan a trip, with adoption highest among Millennials and Gen Z.

And while over half trust AI to find the best prices and create itineraries, only 35 per cent trust it with visa or immigration requirements.

Founder Adam Ewart admitted it can “take the pain” out of research. However, it makes mistakes, he warned.

“It can only provide answers based on the information it has,” he told nine.com.au.

“The best way to use AI for travel is to treat it as a starting point, not a replacement for official travel advice.”

Sometimes AI even makes up places entirely.

Last year, dozens of tourists flocked to a Tasmanian town in search of picturesque hot springs they’d seen online, only to learn that they’d been duped by an AI-generated image.

Experts at RMIT University in Melbourne said you need to put your detective’s cap on.

Read more

Strict new Bali visa rules crack down on influencers and content creators

“Spotting a fake isn’t about finding one giveaway – it’s about building a picture,”  Dr TJ Thomson from the School of Media and Communication said.

He told holidaymakers to “look closely for inconsistencies or glitches” and use multiple sources of evidence.

“If you’re not sure, do a reverse image search on Google,” he said.

Internet security company Mcafee has also warned just how much information you are giving away if you post a holiday photo publicly.

Some AI models have a more than 90 per cent accuracy rate at detecting the location a photo was taken based on the visuals in the photo alone, it said.

And that makes it easy for scammers.

“AI removes the guesswork, allowing attackers to build highly specific, contextual scams at scale,” it warned.

That could include sending believable context around an attack, such as naming the city or country where the victim had been, fooling them into thinking a message or email is legitimate.