Czech mobile operators block millions of scam calls a month

Roughly one in four people in Czechia has already come across scam content made with artificial intelligence, while telecom operators block millions of suspicious calls. According to an Ipsos survey for Vodafone, it is not only individuals who face deception attempts, but companies too.

Scammers need only one photo and 15 seconds of a voice to create a digital clone of a person. That is how Matyáš Boháček, a Stanford University researcher in AI and deepfake technologies and co-founder of Synhawk, explains it.

How deepfake scammers work

Deepfake technology uses machine learning to create or alter video and audio so that they look and sound convincing: a person seems to say and do things they never said or did. Such clips can be used to extract money for fake investments or to damage politicians’ reputations during elections. Scammers use them in SMS messages, calls, voice messages and documents that they send out across various platforms.

According to the survey, 31% of companies have faced identity impersonation. 22% have recorded fake invoices and documents created or altered by AI, and 9% have seen suspicious voice messages or calls made in the name of an employee or company management. Most often these are requests for a payment or a change of bank account. Boháček points out that a photo of almost any adult can be found online, while criminals obtain a voice simply: they call from an unknown number, wait for the person to answer, say hello and ask who is speaking, and then record the conversation. That is why 73% of companies would consider an operator service for checking deepfakes, primarily for invoices and payment requests.

What the operators offer

On 22 October, Vodafone plans to launch Vodafone Verify, an app for checking suspicious photos, videos, voice messages and documents. Users upload a file or a link, and the service analyses the risks and returns a probability assessment with an explanation. The app will be available only to Vodafone customers. Plans include Verify Meetings, which recognises deepfake participants in online meetings, and Verify Network, protection directly in the operator’s network. Vodafone product manager Petr Rogáček stresses: “This is not a magic shield that will always protect you a hundred percent. The model assesses probability and urges caution.”

O2 has “AI Granny Alena” on WhatsApp, available free of charge to everyone at 720 002 002. You can send her a suspicious SMS, link, image or voice message, and she will help you understand what the fraudster is after and how to act. You do not have to be an O2 customer to consult her. According to O2 security director Radek Šichtanc, over the past six months the operator has blocked more than six million attempted fake calls, with a record of 214,000 attempts in a single day in July. Customers send about 3,400 suspicious SMS messages to the number 7726 every day, and on average 20 new fraudulent domains are blocked daily.

T-Mobile, according to spokeswoman Patricie Šedivá, blocks around one million calls a month with its own firewall alone, plus thousands of scam SMS messages a day. You can report a scam with a button on your phone or by forwarding the SMS to 7726. Together with CZ.NIC, the operator also blocks fraudulent websites. As early as 2023, T-Mobile introduced protection against number spoofing using a phone firewall.

No technical measure catches every scam, Šichtanc notes: fraudsters constantly change their schemes. That is why it is important to combine network protection, customer vigilance and education, and operators and banks complement each other: a bank can stop a suspicious transaction, while an operator can intercept a call, a message or a dangerous website.

Read also: Savings accounts in Czechia: bank rates in 2026

Source: seznamzpravy.cz