Yandex Moves Ahead of Scammers with Predictive AI to Combat Ad Fraud

Reducing Losses from Fraudulent Activity

Yandex says the upgrades reinforce the security of its advertising network for all market participants. The new system materially improves the safety of the online environment. User risk has dropped, with clicks leading to phishing sites declining multiple times over. The platform also helps reduce business losses tied to click fraud, a global problem that costs advertisers billions of dollars annually.

Deploying advanced AI-based antifraud capabilities on one of the country’s largest ad platforms affects millions of ad impressions and user interactions. The benefits are distributed across the ecosystem. Russian advertisers see lower exposure to fraudulent schemes. Users face a reduced risk of landing on phishing pages. Yandex strengthens trust in its advertising products, including Yandex.Direct.

Yandex’s initiative aligns with global trends. Major Western platforms also rely on AI to cut fraudulent clicks by double-digit percentages.

Broader Implications

The predictive antifraud algorithms and architectural approach developed by Yandex could serve as a technological case study for other global platforms.

Fraudsters are turning to more complex schemes, which creates a challenging environment for defenders. To track these activities, we are effectively forced to continuously refine detection mechanisms. Machine learning, however, adapts to change far more effectively than static expert rules

Vladislav Balaev

Head of Big Data and Artificial Intelligence Competence Center Practice at LANIT

Within Russia, the company’s AI tools can be scaled beyond advertising to combat fraud across other e-commerce segments and big data processing environments. Planned integration with additional Yandex advertising products and partner programs is expected to increase the share of verified ads and further build advertiser trust while reducing financial exposure. Over time, the technology will require adaptation to evolving attacker tactics. Yandex engineers indicate they are prepared for that escalation.

The Growing Role of Predictive Analytics

Western companies have long deployed AI to address native advertising fraud. Since late 2023, Google has used Gemini AI to detect advertising fraud and invalid traffic across web and mobile platforms. Gemini evaluates data associated with each click, identifying anomalies such as accidental clicks and concealed ads. In a pilot program running from December 2023 through October 2024, invalid mobile traffic declined by 40%.

In 2025, Yandex introduced its AI system for advertising fraud detection. The technology forecasts the behavior of malicious advertisers and blocks dangerous ads before content is swapped for phishing payloads. The system relies on algorithms that simultaneously analyze advertiser behavior and materials published on their websites. By the end of 2025, more than 120,000 accounts belonging to fraudulent advertisers had been blocked.

Enterprise services built on AI increasingly use predictive analytics to prevent fraud in banking, e-commerce and advertising networks. For example, Kount provides protection for online retailers and payment portals by analyzing the risk level of each transaction and issuing real-time action recommendations.

The Future of Antifraud Technologies

The new AI-driven analytics system marks a shift from reactive defense to predictive fraud prevention. It strengthens Yandex’s position as a platform offering advanced security and user protection technologies. Fewer phishing redirects not only enhance ecosystem-wide trust but also reduce operational and financial risk for businesses and end users.

Looking ahead, the technology is expected to evolve toward preventing fraud before it materializes, rather than merely blocking it at the moment of detection. New AI tools for advertisers could provide risk forecasting, projections of secure advertising costs and automated alerts. Cases like this, implemented in Russia, support the competitiveness of the national IT sector in cybersecurity and automated risk management.