Sber Uses AI to Check the Reputation of 300,000 Counterparties Every Day
Sber conducts over 300,000 counterparty checks each day using AI-driven models. Previously, manual verification could take up to five working days, with an annual volume of around 250,000 cases. Today, a comparable volume is processed in a single day.
As financial institutions adopt AI in counterparty screening, the sector is shifting from pilot deployments to full-scale use of technology in core processes – including fraud detection, risk management, and internal controls. It reflects growing demand for domestic machine learning solutions, risk analytics tools, and deep integration with data sources.
For users, the benefit is straightforward: lower exposure to fraud and more reliable banking services. At the national level, the effect is greater transparency in the business environment and continued development of the domestic technology stack. Export prospects for such solutions depend on external constraints and trust in Russian-developed systems.
Shift Toward Automated Monitoring
In the coming years, similar AI systems are likely to move beyond banking into sectors where large-scale counterparty verification is essential, including telecommunications, public procurement, retail, industry, logistics, and the SMB segment.
For companies, this marks a transition from one-off manual checks before transactions to continuous, automated monitoring of partners.
For Sber, this approach lays the groundwork for expanding partner verification services and packaging AI-driven scoring both as an internal capability and as a B2B offering. Export potential is most realistic in the form of platform-based solutions targeting markets in the EAEU, CIS, BRICS, and other aligned jurisdictions. However, scale will depend on access to local data and regulatory requirements.
Toward Comprehensive Risk Scoring
In 2020, Sber patented an AI system designed to assess the legal capacity of corporate entities, signaling a long-term strategy to automate company verification as a standalone technology domain. Against this backdrop, the role of government digital services is also increasing. In 2024, more than 652 million queries were submitted to the “Prozrachnyy biznes (Transparent Business)” service, underscoring strong demand for counterparty data.
Meanwhile, private-sector players are deploying automated scoring systems and large-scale client and supplier screening. Integration with 1C platforms enables full coverage of counterparties while reducing manual workload for security teams to about 10% of operations. This is illustrated by the Remtekhkomplekt case: integration with 1C allowed 90% of counterparties to be approved automatically, with only 10% escalated for manual review.
By 2024–2025, commercial risk analytics systems had effectively become standard practice for businesses. Companies are moving beyond basic registry checks to comprehensive risk scoring that combines sanctions exposure, media coverage, financial indicators, and business reputation. The Bank of Russia reports expanding use of AI across the financial sector – from fraud prevention and identity verification to analytics, forecasting, and risk management. Sber’s 300,000 daily checks fit squarely within this broader trend.
Industrial-Scale AI in Corporate Security
The widespread use of AI models in the financial sector points to a transition toward industrial-scale deployment in corporate security. Automation now covers hundreds of thousands of operations daily, extending well beyond isolated pilot projects. This indicates the maturity of industrial AI at the intersection of fintech, cybersecurity, compliance, and big data.
Over the next two years, AI-based counterparty scoring is expected to integrate more deeply into ERP systems, 1C platforms, banking interfaces, and electronic document management systems, while manual review will remain for complex cases. Over a three- to five-year horizon, the market is likely to move toward continuous monitoring and the emergence of AI agents capable not only of identifying risks but also recommending response strategies.
