Yandex has started rolling out Universal Anti-Fraud, a unified AI platform intended to combat bots, ticket manipulation, and other forms of digital fraud across its ecosystem.
The new system allows protection to be connected to any service in just two to four days, a sharp reduction from the several months previously required to develop individual solutions for each product.
Dozens of Yandex services are already integrated, including Eda, Afisha, Puteshestviya, and applications featuring the Alice AI assistant.
Although attack scenarios differ by service, the core objective remains the same: fraudsters attempt to deceive algorithms for financial or other gain.
In Afisha, the platform detects bots that mass-book tickets for popular events, enabling scalpers to manufacture artificial scarcity before reselling the tickets at inflated prices on secondary markets.
In Eda, the anti-fraud module reviews support requests and has become more than ten times more effective at identifying and blocking repeated fraudulent complaints about the same order aimed at securing compensation.
The underlying technology combines neural networks, analytical methods, and over one hundred rules describing major attack classes. The platform processes traffic in real time and tunes protection parameters individually for each connected service.
Its principal strength lies in centralized updates: when Yandex identifies a new fraud scheme, the corresponding defenses are rapidly propagated to every linked product, allowing the system to stay ahead of evolving threats.
