Cybercrime in the Czech Republic: damages exceeding 1.9 billion CZK in seven months, AI a weapon for both sides
In the first seven months of 2026, losses from fraud at Czech banks rose by 79 per cent to 1.9 billion crowns. Perpetrators and banks alike are using artificial intelligence.
Cybercrime in the Czech Republic growing, AI helping both banks and fraudsters
In the first seven months of 2026, fraudsters inflicted damages of 1.9 billion crowns on clients of Czech banks — a year-on-year increase of 79 per cent, according to investujeme.cz. In total, perpetrators attempted to defraud clients of more than ten billion crowns. Bank systems and bank employees managed to protect over 8.1 billion crowns, or 81 per cent of the threatened sum.
The number of recorded attacks rose by approximately one-fifth to 62,705 cases. The average loss per affected client increased by nearly ten thousand crowns, reaching 30,335 crowns.
Social engineering as the principal weapon
Bank technological security is functioning — fraudsters therefore increasingly target clients directly. If they cannot circumvent bank systems, they attempt to persuade the victim to send money themselves.
"Banks are succeeding in halting the vast majority of suspected transactions thanks to advanced security systems and have been investing billions of crowns in client protection over the long term. However, fraudsters are increasingly betting on social engineering. They create time pressure, provoke fear and impersonate bankers, police officers or investment advisers. Their aim is to convince the victim to confirm the payment themselves, hand over sensitive information or transfer money to a purportedly secure account," says Aleš Blažek, president of the Czech Banking Association and chief executive officer of the ČSOB group.
Banks deploying AI and behavioural biometrics
Financial institutions are using artificial intelligence and machine learning primarily for fraud detection in real time. Systems monitor transactions and compare them with the typical behaviour of a specific client. Suspicion can be raised by an unusually high payment, login from an unknown device or withdrawal abroad — upon detecting any deviation, a bank may request additional verification or temporarily halt the payment.
Behavioural biometrics, which evaluates writing style, mouse movement or the way a telephone is held, also contributes to protection. Some banks offer the possibility of verifying directly in the application whether the caller genuinely works for the institution.
Results have been recorded by Visa as well. Thanks to a combination of artificial intelligence, data and multi-layered protection, Visa has managed to reduce the number of frauds in card payments in Europe by 24 per cent over the last three years. Since the end of 2024, Visa has stopped more than 25,000 fraudulent merchants. Its Cyber Fusion Centre deals with approximately one hundred incidents daily, with AI contributing to the resolution of 80 per cent of them.
Generative AI reducing costs of fraud preparation
The development of generative artificial intelligence simultaneously makes the work of perpetrators easier. In a short space of time they can create a linguistically flawless message, imitate the communication style of a trustworthy institution or prepare false image and audiovisual content. Fraud is therefore not revealed by poor Czech or amateur graphics, which people previously relied upon to identify it.
A Visa survey showed that 61 per cent of people mistook content created by artificial intelligence for genuine content. People who cannot recognise AI-generated content, according to the survey, become targets and victims of fraud almost five times more frequently.
"Frauds are becoming increasingly sophisticated and with the development of artificial intelligence it is harder for people to discern what they can trust. Our survey showed that 61 per cent of people confused content created by AI with genuine content, and precisely these people become targets and victims of fraud almost five times more frequently. This is why technology must keep pace alongside consumer caution," warns František Jungr, director of the product division for Visa in Central and Eastern Europe.
Fraud changing shopping behaviour of millions of Czechs
The impact of fraud does not end with direct financial loss. According to a Visa survey, approximately 2.2 million Czechs curtailed or completely stopped shopping online after becoming targets of fraud. People who encountered fraud subsequently reduced their number of card payments by roughly one-fifth. Cybercrime thus weakens confidence in the digital economy alongside direct financial losses.
45,000 fraudulent advertisements on Facebook and Instagram
Social networks play a significant role in reaching potential victims. The Centre for Online Risk Research recorded more than 45,000 fraudulent advertisements on Facebook and Instagram from the beginning of the year to mid-August 2026. In the Czech Republic they received approximately 144 million impressions.
Advertisements frequently exploit well-known personalities, media outlets and established brands — referring for example to false investment platforms or pages mimicking the appearance of trustworthy news websites. Perpetrators use edited photographs and deepfake videos in which personalities seemingly recommend a specific investment.
"Fraudsters often use deepfake videos or images edited with the help of AI. Advertisements typically refer to false pages mimicking the appearance of trustworthy media, companies or institutions. It often suffices to use a well-known name, a trustworthy brand or a current event and place it in an advertisement that looks like a standard offer," says Vendula Prokůpková from the Centre for Online Risk Research.
Source: Google News CZ — Crime (cs)