03.09.2025 15:39:00
Дата публикации
In Kazakhstan banks, artificial intelligence is starting to play a key role in the fight against financial fraud. This is a response to the growth of cybercrime, the number of which has increased by 20% in recent years.
New AI systems can analyze customer behavior and identify suspicious transactions in real time. They are integrated into the existing infrastructure of banks without radical changes, which makes implementation fast and efficient.
The national regulator has already obliged banks to use specialized software. Two types of solutions are currently used: transactional antifraud monitors card transactions, and session antifraud monitors user actions in online banking.
However, these systems have limitations: they do not see the whole picture if the client uses several services at the same time. This creates vulnerabilities that can be exploited by attackers.
AI offers a more flexible approach. It analyzes large amounts of data, identifying hidden connections and anomalies. Hybrid models that work simultaneously at the transaction and user session levels are especially effective.
If the system detects deviations from the client's usual behavior, the transaction is automatically categorized as suspicious. Depending on the risk, the bank may request additional verification or temporarily block the transfer.
The main advantage of AI is its ability to learn. Algorithms adapt to new fraud schemes, and updates are implemented into the system in a matter of days, not months, as before.
But this can also create certain risks. Banks must have very careful human control over AI activities, as well as to ensure that AI solutions do not harm clients by mistake.
However, the introduction of AI into information security systems is not just a technological trend, but a necessary step to increase the stability of the financial system, experts believe.
(the text is translated automatically)
New AI systems can analyze customer behavior and identify suspicious transactions in real time. They are integrated into the existing infrastructure of banks without radical changes, which makes implementation fast and efficient.
The national regulator has already obliged banks to use specialized software. Two types of solutions are currently used: transactional antifraud monitors card transactions, and session antifraud monitors user actions in online banking.
However, these systems have limitations: they do not see the whole picture if the client uses several services at the same time. This creates vulnerabilities that can be exploited by attackers.
AI offers a more flexible approach. It analyzes large amounts of data, identifying hidden connections and anomalies. Hybrid models that work simultaneously at the transaction and user session levels are especially effective.
If the system detects deviations from the client's usual behavior, the transaction is automatically categorized as suspicious. Depending on the risk, the bank may request additional verification or temporarily block the transfer.
The main advantage of AI is its ability to learn. Algorithms adapt to new fraud schemes, and updates are implemented into the system in a matter of days, not months, as before.
But this can also create certain risks. Banks must have very careful human control over AI activities, as well as to ensure that AI solutions do not harm clients by mistake.
However, the introduction of AI into information security systems is not just a technological trend, but a necessary step to increase the stability of the financial system, experts believe.
