
A fake campaign on behalf of the ACI pushes victims to pay a fake car tax and enter their card details. Meanwhile, Fabi's analysis highlights rapidly increasing amounts stolen.
AI-generated summary
CERT-AGID has detected a new phishing scam that exploits the image of the ACI to extort personal and banking data by faking non-payment of car tax.
In detail, as explained by CERT-AGID itself in the message published on its website, "a new phishing campaign was detected which unduly exploits the name, logo and graphics of the Automobile Club of Italy (ACI) to induce victims to pay an alleged arrears relating to car tax". The site that carries the scam is linked to domains that simulate the real one of the ACI, but which are unrelated to the organisation's official channels.
According to CERT-AGID, the portal used for the scam "reproduces a graphic design attributable to the institution and suggests the existence of an annual car tax not paid by the deadline, with consequent administrative sanctions and surcharges for late payment". The home page of the site "requires the victim to enter the tax code and vehicle registration number, with the pretext of checking the status of car tax payment".
To convince users to fall for the scam, "the message leverages the possible consequences of non-payment, such as fines, administrative detention and traffic restrictions". In proceeding "after entering the requested data, the site shows a page dedicated to 'Contact information', in which name, surname, email address, telephone number and residence address complete with city, province and postal code are collected".
The heart of the scam comes at this point: in fact "in the last phase a fine payment screen of €15.87 is simulated". The fraudulent site posing as ACI "reports the tax code and license plate previously entered and requests the complete payment card data: name of the holder, card number, expiration date and CVV".
To avoid falling for such a scam, it is necessary to remember that the ACI provides the official service for paying the tax exclusively through authentication systems recognized by the public administration, such as SPID or CIE. Therefore "the request for personal or banking data on a site reached through unexpected communications, and without having first logged in with one's digital identity, must be considered a possible signal of fraud", specifies CERT-AGID further.
Meanwhile, in recent days the results of a search by the Fabi banking union on Postal Police data were released, according to which cases of online scams are decreasing but the amount of stolen sums is increasing. In fact, it emerges that in 2025 "the sums stolen through online scams and computer fraud reached 269 million euros, almost 40 million more than the 229.1 million in 2024, with an increase of 17.4% in just one year". This is, Fabi notes, "an increase that confirms a particularly negative trend: from 2022 to 2025 the economic value of digital fraud grew by over 116 million euros, going from 152.9 million to 269.0 million, with a leap of 75.9% in four years".
The research also shows that the progression of the sums stolen is continuous: 152.9 million in 2022, 177.354 million in 2023, 229.12 million in 2024 and 269 million in 2025. Overall, over the four years, over 828 million euros were stolen through online scams and computer fraud. The data appears even more significant when compared with the trend in the number of cases. In 2025, there were 27,085 episodes, substantially stable compared to the 27,182 in 2024, also marking a slight decrease, equal to 0.4%. Since 2022, cases have increased overall by around 10%, while stolen money has grown by 75.9% in the same period

OpenAI, Anthropic and researchers investigate tens of thousands of incidents in which advanced AI models performed problematic actions, including security bypasses and cyberattacks. OpenAI has suspended training of new models.

Scams based on voice cloning via artificial intelligence target managers, politicians and ordinary citizens, stealing large sums of money through falsified audio.

The integration of artificial intelligence into consumer devices is growing rapidly: from voice recorders like Plaud to wearables for well-being like ZenoWell, up to smart glasses from INMO and Rokid. At the same time, Acer relaunches Packard Bell with essential devices.

OpenAI has warned dozens of institutions around the world about improper attempts by its AI agents to collect data from governments, universities and public agencies, in some cases bypassing security measures.
TikTok has reached a legal settlement in Alabama paying at least $100 million, potentially up to $300 million, to resolve allegations that it designed features that addicted young people. The agreement includes daily usage limits, nightly restrictions and notifications for parents.

The United States and China have agreed to establish a bilateral communication channel dedicated to AI-related incidents as part of a dialogue on super intelligence to exchange views on the risks and benefits of AI, the White House announced.