Russian telecom operator MTS has carried out a deep upgrade of its anti-fraud systems, integrating three specialized AI models into the "Zashchitnik" service at once. The result is impressive: response time to a suspicious call has been halved — from 30 to 15 seconds, and the accuracy of detecting complex multi-stage fraud schemes has tripled.
This is not a cosmetic improvement, but a fundamentally new level of protection. It is based on comprehensive analysis that takes place in real time. Every second, the system evaluates more than 1,100 call parameters, analyzing not only the technical characteristics of the connection, but also the context of the conversation, the speed of the interlocutor's speech, and other behavioral markers indicating a potential threat.
Hunting multi-stage attacks
Particular emphasis in the update is placed on combating multi-stage fraudulent scenarios, where attackers are not limited to a single call but build an entire chain of psychological pressure on the victim. To improve effectiveness, the models were retrained on new data and supplemented with more complex language algorithms, allowing them to recognize even the most sophisticated manipulation tactics.
From January to July 2026, the service managed to detect and block more than 1.5 billion spam calls and fraudster calls. This figure clearly demonstrates the scale of the threat facing Russian subscribers. It is important to understand that phone calling is only the first link in the attack chain. It is often followed by the transfer of funds, including in cryptocurrency, which makes such crimes especially difficult to investigate.
In parallel with protective measures, there is also a rise in criminal activity using AI. Attackers use chatbots to "poison data" and create fake stores, and also actively use fictitious cryptocurrency access, luring victims to phishing sites disguised as closed gateways of the Moscow Exchange. Special mention should be made of drainers — malicious programs disguised as partner investment programs that empty wallets after connection via QR code. Creators of infostealers on Telegram are also keeping pace, stealing passwords and crypto wallet data under the guise of secretary bots.
As an analyst, I see in this MTS update an important signal for the entire industry: the fight against fraud is moving to a fundamentally new technological level. However, it is worth remembering that AI is a double-edged sword. While some companies use it for protection, others use it for attacks. The key factor becomes the speed of adaptation and the quality of the data on which algorithms are trained. Telecoms investing in such solutions gain a significant competitive advantage in subscriber security matters.