A WhatsApp message purporting to be from the CEO of Intesa Sanpaolo, followed by a phone call imitating the voice of a lawyer: the Italian private bank lost 95 million euros, of which 53 million have been recovered.
The case illustrates the shift of voice deepfakes from theoretical risk to very real harm. Paolo Molesini, president of Fideuram, the private bank of the Intesa Sanpaolo Group, first received a WhatsApp message introducing itself as coming from the group's CEO, Carlo Messina, who was asking for his help with an urgent operation abroad. A call followed, in which the fraudsters used an AI-generated voice mimicking that of a senior lawyer at a firm linked to the bank.
Convinced of the legitimacy of the request, the executive arranged transfers for a total of 95 million euros (about $108 million), mainly to accounts in China and Hong Kong. The authorities were able to recover nearly 53 million; a significant part of the rest was converted into cryptocurrencies and remains untraceable. The Milan prosecutor's office has opened an investigation for computer fraud against a foreign national residing outside Europe.
A threat estimated in the tens of billions
The publisher IdentifAI, which specialises in the detection of synthetic content, estimates that generative AI could increase fraud-related losses in the United States to $40 billion by 2027. According to its data, X accounts for 42.1% of the deepfake frauds recorded, ahead of LinkedIn (14.5%).
Read the full article on the original source
Go to the source →