Yandex Uses AI to Boost Call Filter Accuracy by 40 Percent

With new neural networks analyzing suspicious patterns, Yandex’s caller ID system is faster and more precise in detecting scam calls.

Russian tech giant Yandex has upgraded its caller ID service with neural network technology to fight phone scams. The new system is now 40 percent more accurate in flagging suspicious calls, and it can analyze new phone numbers in just 1.5 minutes—a major improvement from the previous 10-minute average, according to CNews.

The AI-enhanced system evaluates over 300 variables, such as call frequency, duration, and caller behavior. For example, if a number is dialing dozens of people in a short span and ends the calls after just a few seconds, it's flagged as potentially dangerous. Importantly, all data is processed anonymously, without using any personal user information.

Scammers often rely on fresh numbers with no call history, impersonating couriers, tax officials, or bank employees to trick victims into giving up SMS codes or personal data. The updated Yandex filter is designed to spot these patterns quickly and alert users in real time.

The neural networks significantly enhance the service’s effectiveness: faster identification of threats, better filtering of unwanted calls, and protection against robocalls and emerging fraud schemes.

Yandex’s AI-powered move makes mobile communications safer, saving time and stress for millions of Russians. As artificial intelligence continues to advance, it’s quietly reshaping digital safety and reliability.