Scammers used an AI voice clone and a fake WhatsApp message to move €95 million out of an Italian bank

Alfonso Maruccia

Posts: 2,763   +1,080
Staff
AI Scammed: As AI models grow in complexity and capabilities, scammers are expanding their business opportunities as well. Social engineering techniques have fully translated into this new wild AI world, and high profile executives are losing their jobs as a consequence.

Sources familiar with the matter have unveiled yet another case involving an executive who lost his high-paying position due to an AI-based scam.

Paolo Molesini, chairman of private banking institution Fideuram, was targeted in a complex fraud scheme that began in February 2026. Molesini resigned from his position soon afterward, and the bank is still trying to recover the full amount extracted by cybercriminals.

Fideuram is the private banking arm of Intesa Sanpaolo, Italy's largest bank and one of Europe's biggest banking institutions. Like other private banks, the organization provides services to high-profile customers with significant financial assets. The sources said that Molesini first received a WhatsApp message that appeared to come from Intesa Sanpaolo CEO Carlo Messina.

Messina was apparently asking Molesini to help with a large overseas transaction. After the initial message, Fideuram's chairman received a phone call from a senior partner at a major law firm. However, there were no humans behind the call: The sources confirmed that cybercriminals used AI models to impersonate the lawyer's voice, convincing Molesini that the initial request for help had indeed come from his "boss," Messina, at Intesa Sanpaolo.

Fideuram was instructed to proceed with the transaction. The bank transferred a total of €95 million (around $108 million) to a series of accounts in foreign countries, including China and Hong Kong. However, Fideuram's team detected anomalies in the transactions and alerted the banks where the money had been transferred.

Thanks to local authorities and banking institutions in Italy, Portugal, and China, Fideuram was eventually able to recover €53 million from the fraudulent transactions. The remaining sum was apparently lost after passing through a complex network of overseas accounts, ultimately being converted into an undisclosed cryptocurrency.

Molesini resigned from his position in March without providing a specific explanation for his decision. The newly unveiled incident might explain why the executive left his high-paying position at Fideuram. As a private bank, the institution provides its affluent customers with a private broker to personally manage their financial needs.

The Fideuram incident highlights how AI models have become a significant security threat and a potential financial disaster in the making. Milan-based prosecutors are now trying to identify the alleged cybercriminals behind the AI-generated heist. One person living outside the European Union is reportedly under investigation for computer fraud.

Permalink to story:

 
What a dumba55 ... would have taken 30 seconds on a second phone to call the big boss to confirm, while still on the line with the fraudsters.

 
How is it possible to transfer millions after a phone call, without 2 factor authentication and other measures?

It's just plain stupidity, AI has nothing to do with it.
I mean, even if it were the real owner of the money, tomorrow he may claim he didn't request the transfer and someone imitated his voice. There are people who can do that, not to mention that relatives often sound very similar, especially over a phone.
 
Here is the real reason why you are allowed to access AI:


Digital IDs are coming, on purpose. Problem. Reaction. Solution. Don't say you weren't warned.
 
The scary part isn't that AI can clone a voice. It's that €95 million could apparently be moved after a WhatsApp message and a convincing phone call. At that scale, the authorization process seems like the bigger vulnerability.
 
We are fast approaching a digital "event horizon", where AI impersonation becomes so ubiquitous, that conducting business online becomes financially unviable. Too many opportunities for exploitation and two few legitimate use-cases. Of course, people want to do what is easiest, but if every door is just " oops, all scams", then people will assume deception by default. Or, they should.

I mean, it used to be common sense that you "don't believe everything you see online". What happened to that? I guess Facebook happened and then TikTok followed. Really feels like we're cooked, we just don't know it yet...
 
Back