Skip to content
Submit incident
AII-1001-86799
ReportedReported by fewer than three independent publishers, with no official finding.Supporting sources are pending verification.

ChatGPT visa hallucination strands Australian author at airport

Mark Pollard, an Australian author, was stranded at an airport after ChatGPT incorrectly told him he did not need a visa for Chile. He missed a conference he was scheduled to present at, and when he shared his experience on social media, the video accumulated over 1.5 million views.

22 March 2025 Β· OpenAI Β· Reliability and hallucination
22 March 2025
AII-1001-86799
Quick facts
Category
Reliability and hallucination
Country
Australia
Date
22 March 2025
Drafted by
Raidu Incident Detection

What happened

Mark Pollard, an Australian author specializing in marketing strategy, asked ChatGPT whether he needed a visa for a short stay in Chile, where he was scheduled to deliver a conference.

ChatGPT stated that no visa was required, but this information was incorrect. Upon arriving at the airport, Pollard discovered he needed an electronic visa obtained in advance from a Chilean consulate.

Unable to board the aircraft without the required visa, Pollard missed his conference and shared his experience on social media on 22 March, accumulating over 1.5 million views on Instagram.

Reported impact

Pollard was prevented from entering Chile and missed the conference he was scheduled to present at.

The social media post detailing his experience went viral, drawing millions of views and mixed reactions, including mockery, with some people questioning whether the account was genuine.

Index analysis

When researchers from Tech&Co posed the same question to ChatGPT, the system repeated its erroneous claim that no visa was required for an Australian short stay in Chile.

ChatGPT only corrected its answer when asked directly for its source, then acknowledged the error and provided accurate information about the electronic visa requirement in place since 9 May 2020.

Governance controls

OpenAI includes a disclaimer on ChatGPT's interface stating that 'ChatGPT can make errors' and advising users to 'verify important information.'

Pollard acknowledged he should have verified the information using a search engine, noting that he chose ChatGPT because he found contemporary search results cluttered with sponsored content.

Classification

Category
Reliability and hallucination
Industry
Travel, leisure and hospitality
Country
Australia
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Economic or property, Public interest, Reputational
Harmed
Consumers
Failure modes
Agent misaction, Hallucination, Overreliance
AI systems
ChatGPT

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident