Broke? No way! Google AI invents insolvency and almost ruins a company

When the machine lies: How an AI plunged Seyvillas into chaos

Imagine this: Your company has been running successfully for years, customers are booking, the ratings are top - and suddenly you read online that you are allegedly bankrupt. This is exactly what happened to the Seychelles tour operator Seyvillas. And the "culprit"? Not a tabloid newspaper, not a begrudging blogger - but Google's artificial intelligence.

In a so-called "info box" - directly and prominently above the search results - there was suddenly the claim that Seyvillas was insolvent. No source, no citation, no reality. Simply a made-up "hallucination" of the AI, as the jargon goes.

False information with real damage

Managing Director Julian Grupp was shocked. And rightly so: the fake bankruptcy spread quickly. Customers canceled bookings or contacted the company with concerns. The damage to the company's image was enormous. It was only after Seyvillas took legal action and called in a lawyer that Google reacted and removed the false content. But trust had already been damaged - among customers and business partners.

And that's not all: Seyvillas is considering further legal action, including an injunction against Google. Because one thing is clear: such mistakes should not be allowed to happen - especially not in search engines that many users regard as "truth".

AI can do a lot - but it can also break a lot

The problem is not new: so-called "AI hallucinations" are a well-known phenomenon. Artificial intelligence invents content that sounds plausible but is completely wrong. In harmless cases, this makes people laugh - in the worst cases, as here, it puts entire livelihoods at risk.

It becomes particularly problematic when this content is automatically integrated into infoboxes by large platforms - without editorial control. What remains: A lie that spreads across the web like a virus.

Industries such as tourism, finance and healthcare are particularly susceptible. A fictitious diagnosis, a false insolvency rumor or manipulated company information - and the damage is done. And unlike traditional media, there is no clear person in charge, no "editor-in-chief" who is liable for mistakes. The machine writes - and the person suffers.

Who has to take responsibility?

Those who invent false bankruptcies should not be allowed to hide behind algorithms. It's almost cynical that a billion-dollar corporation like Google sends out an AI that can effectively ruin companies - and then shrouds itself in silence when things go wrong.

The argument "AI just did it wrong" is not a free pass. If digital systems act like humans, then they must also take responsibility like humans - or rather: their operators. Anything else would be a license for disinformation. And we already have more than enough of that.