Artificial Intelligence Enters Preventive Control of the Food Market
The regional offices of Russia’s agricultural watchdog in the Tver and Yaroslavl regions are using neural networks to analyze vast datasets, identify early signs of potential violations, and organize targeted inspections.
Digitizing Risk Signals
Automation of inspection and enforcement activities in Russia’s agro-industrial sector began as early as 2007. Today, the regulator operates 10 information systems. Artificial intelligence was introduced three years ago, as neural networks matured and demonstrated growing analytical capabilities.
AI now helps address one of the sector’s most critical challenges – preventive oversight of agricultural producers and food manufacturers. Instead of mass inspections, authorities apply a risk-based approach, focusing only on enterprises where the probability of violations is high.
Identifying such cases is where AI plays a central role. Digital platforms such as Mercury and Horriot make it possible to see a complete operational picture of a company and trace a product’s entire journey – from field and farm to retail shelf and final sale.
“Artificial intelligence analyzes enormous volumes of data and helps detect potential inconsistencies. In 2025, the system flagged around 65,000 indicators of possible violations, but actual breaches were confirmed in only 4% of cases. Using the Mercury and Horriot platforms, specialists can see the entire production chain – from farm and field to the retail counter,” explains Elena Lapina, head of the regional agricultural watchdog office in the Tver and Yaroslavl regions.
Targeted Oversight
Digital systems allow inspectors to detect inconsistencies in traceability data and identify enterprises that are most likely to violate regulatory requirements.
In 2025, for example, one enterprise in the Tver region – the Bologovsky Dairy Plant – introduced about 25% of its products into circulation with traceability violations. This triggered a comprehensive inspection that uncovered additional issues. As a result, compliant companies are spared unnecessary inspections, while violators receive focused regulatory attention.
“Among the advantages of using AI is its motivational effect on producers. Their operations have become more transparent. The systems help combat food fraud and uncover deceptive schemes – cases where 10 kilograms of saury are somehow ‘turned into’ tons of pork, poultry, or other fish products. Our office shut down 10 so-called phantom livestock production sites, preventing 1,896.23 tons of products of unknown origin from reaching store shelves. Exhaustive measures were taken against all violators,” emphasizes Irina Slutu, deputy head of the regional agricultural watchdog office.
Bringing Land Back Into Use
Oversight extends beyond food products to the resources used in agricultural production, including farmland. The regulator has the authority to forcibly reclaim land from owners who use it improperly. Through these measures, 4,408 hectares of agricultural land have been returned to productive use in the Tver and Yaroslavl regions. Expanding monitored areas is increasingly supported by drones, which since this year are being used for both land-use and phytosanitary supervision.
Strict enforcement against counterfeit products helps create a healthier environment for compliant producers. Over the first 11 months of 2025, meat production in the Tver region rose by 33%, poultry output increased by 14%, and dairy production grew by 24%. Export volumes increased by 18% compared with the same period in 2024.
As phytosanitary requirements for food products become more stringent, full digitalization of control procedures – including the use of AI and drones – is expected to become a competitive advantage for regional producers in foreign markets. Integration of digital oversight systems is also set to accelerate trade flows with partner countries.
