Google AI hallucination falsely declares Japanese company bankrupt, sparking major panic
Business gets buried by bot.
Google glitch. Image by Cybernews.
Follow us
- Google falsely said Japanese textile wholesaler Iwate Sen-i had ceased operations and entered liquidation.
- The error alarmed employees, customers, suppliers, and business partners, despite the company still operating normally.
- The false claim likely came from Google misreading a paywalled local article about another clothing wholesaler.
- The case highlights the business risks of AI summaries when search engines publish unverified claims prominently.
Key Takeaways by nexos.ai, reviewed by Cybernews staff.
A routine Google search wrongly told users that the Japanese textile wholesaler Iwate Sen-i had "ceased operations" and entered special liquidation proceedings.
And with the search result appearing at the very top of the findings on July 10th, an employee spotted the glitch and informed the company's founder.
Iwate-Sen-i is a mid-sized company that employs 90 people and brings in around ¥2 billion ($12 million) in annual sales. The company is operating as expected, yet Google decided that it had run into financial difficulties.
It was a company employee who informed a senior executive via a screenshot, and the alarm had a ripple effect in both internal and external capacities for the Iwate-Sen-i. Within days, the company's phone lines were busy with concerned customers and suppliers, worried that they may be in jeopardy.
And, with brand loyalty and pedigree especially prevalent in Japanese culture, 4 business partners even visited the company's office in person to find out what was happening.
Check if your data was exposed
Find out if your email and related personal information have appeared in known data breaches.
What caused the glitch?
The mistake appears to have stemmed from an article about another clothing wholesaler in the same town that had genuinely entered special liquidation.
As reported by the Asahi Shimbun, Google gave an objectified response, claiming “AI may sometimes misinterpret information on the web or fail to take proper account of its context.”
Google also refused to comment on the specifics of the case.
The ever-scrupulous senior executive Yoshitaka Watabe, however, knew there was a particular reason below the surface.
The Cybernews community is talking about this. Be a part of the conversation.
As Watabe dug deeper, he found that local publication The Iwate Nippo had featured a headline behind a paywall that a clothing wholesaler “Yahaba clothing wholesaler ordered into special liquidation.”
And because the company name was unspecified in the headline, the belief is that the subscriber-only article was not available for AI to scrape, which likely led to conflating it with the textile company Iwate Sen-i.
A local professor of library and information science at the University of Tsukuba believes that Google should provide stronger safeguards against such ruinous claims.
“Google should have a mechanism that withholds the AI summary altogether in such cases,”explained Professor Makota Kato.
On July 19th, Google finally managed to flag the description as "unsubstantiated information,” or in other words, a hallucination.
AI hallucinations have been a major red flag on the tech radar for some time now and are seemingly here to stay.
