MTS Speeds Up Its “Defender” Anti-Fraud Service Sixfold
Multiple AI models now work in tandem to detect and block fraudulent calls, analyzing conversations in real time.
AI on the Front Lines
Phone scammers are constantly inventing new ways to trick people out of their money. Their tactics are often sophisticated enough to fool even cautious users. To counter this, telecom operators worldwide have been upgrading their defenses. In Russia, one of the strongest trends in recent years has been the adoption of artificial intelligence.
MTS, one of the country’s largest operators, has developed its own anti-fraud service called Defender. The company recently rolled out a major update. The new AI models now evaluate every phone number daily against more than 200 fraud indicators — from call frequency and the ratio of inbound versus outbound connections to the percentage of successful completions.
When a subscriber is on the line, three AI models simultaneously assess the conversation, monitoring up to 1,000 parameters in real time and flagging anomalous traffic. The enhanced system now identifies fraudsters with 95.17% accuracy.
Processing Hundreds of Millions of Calls Daily
The upgrade is not just about accuracy — speed has also increased dramatically. Previously, it took Defender around three minutes to confirm a suspicious call. Now, that process takes only 30 seconds, making the system six times faster. The AI currently processes up to 200 million calls per day.
MTS has also reinforced its defenses against aggressive spam calls, including loan offers and financial scams. Big Data specialists retrained Defender’s anti-spam modules, leveraging decomposition techniques, optimized machine learning algorithms, and large language models (LLMs) to analyze speech content. As a result, the system detects and blocks unwanted calls 30% faster than before.
Andrey Biichuk, Product Director of Defender, explained: “Our goal is to create a digital environment where subscribers feel as protected as possible. Detecting and blocking scam calls in just 30 seconds is a breakthrough made possible by our in-house AI development.”
The results are tangible. In the first half of 2025 alone, Defender blocked 1.5 billion fraudulent calls — a 6% increase compared with the same period in 2024.
A National-Level Security Challenge
Russia’s fight against phone scams has moved well beyond corporate responsibility and into the realm of national security. Since late 2022, the federal regulator Roskomnadzor has operated a system known as Anti-Fraud, to which all major carriers are connected.
Under this framework, data on every call is routed to a central system. Operators query the database in real time to confirm whether a call is legitimate. If validated, the call proceeds. If not, the operator blocks the connection. By 2024, more than 1,162 carriers had joined the system, covering 99.6% of all phone numbers in Russia.
Carriers are also innovating on their own. MTS launched Defender into full-scale production in late 2024, while other companies — including T-Mobile Russia, VimpelCom, and SberMobile — use a different strategy, preemptively blocking suspicious calls before they reach subscribers.
The Road Ahead
Security services are evolving along several critical lines: boosting detection speed and accuracy, and expanding protection to messenger-based communications — a growing vector for fraud.
Although initially designed for Russia, technologies like Defender carry export potential. Neighboring CIS countries, Eastern Europe, and Asia face the same global challenge of telecom fraud and may find value in adopting such systems.
What stands out is the level of sophistication Russian engineers have reached. No technology can guarantee absolute safety, but minimizing incidents to the lowest possible level is increasingly achievable.
