AI Helps Uncover First Cartel Case in Russia’s Public Procurement System
As Russia’s Federal Antimonopoly Service reported, an antitrust case involving bid rigging in public procurement has been opened for the first time based on findings generated by artificial intelligence. The Antikartel State Information System, which uses AI-powered analytics, has demonstrated its effectiveness.
The Antikartel State Information System analyzed tenders involving Stroyprofigrupp LLC and MT Group LLC. From 2024 to 2026, the companies allegedly created the appearance of competition in bids for building maintenance services at children’s educational institutions in Odintsovo, Moscow Region. The contracts totaled 676.6 million rubles (approximately $8.3 million). After the Antikartel system flagged the violations, a cartel case was opened. Investigative proceedings are now underway.
For the first time, an antitrust case was opened not in response to a competitor’s complaint or as a result of a routine inspection. This time, artificial intelligence flagged the potential violations. The neural network analyzes information about procurements, bidders, price offers and tender results, looking for even subtle signs that participants in public procurement may be coordinating their actions. These include matching IP addresses used to submit bids, similar document templates and identical errors. The system can also detect patterns in which the same bidders take turns winning similar tenders, as well as coordinated price changes.
The agency trains the neural network using previously identified violations. By fall 2025, developers had uploaded data on 9.2 million procurement procedures to the Antikartel system. The AI does not initiate cases itself. It flags potential violations, which are then reviewed by antitrust agency staff. The better Antikartel performs, the stronger oversight of public procurement becomes and the more effectively public funds can be spent. The benefits extend beyond the government: businesses also gain from more transparent procedures and fairer competition.
From Russia to BRICS
After its launch, the Antikartel system was rolled out across every region of Russia. This gave the agency digital oversight of public procurement at all levels. In effect, the country is preparing for a fundamental overhaul of its oversight model in the coming years. Previously, the Federal Antimonopoly Service (FAS) responded to reports of violations and conducted scheduled inspections. It now has a tool for monitoring every tender on a daily basis.
At an October 2026 meeting of the BRICS+ Working Group on Cartels, held as part of the International Procurement Conference, FAS reported that the Antikartel system already contained data on 28.6 million tenders. It had classified more than 91,000 of them as high- or medium-risk.
Representatives of the antitrust authorities of Egypt and Malaysia showed strong interest in how the system works. This suggests that Russia’s experience using AI to detect violations of antitrust law could find demand internationally.
From the “Big Digital Cat” to an AI-Powered System
FAS spent several years preparing to use artificial intelligence on a large scale to analyze public procurement. In 2021, the agency launched the Bolshoi Tsifrovoi Kot (Big Digital Cat) project, a screening tool that analyzed data in real time and helped identify suspicious signs of bid rigging. In 2022, FAS added the development of the Antikartel Automated Information System to its digital transformation program. In 2024, the antitrust agency began developing a new state information system. In 2025, the Antikartel system received a legal framework through the adoption of a federal law and a government resolution defining its objectives, functions and operating principles. The system went live on August 1, 2025. By 2026, machine-learning algorithms had been fully integrated into it.
The first major cartel case opened on the basis of findings from AI and the Antikartel system demonstrates that the government now has an effective tool for identifying risks in public procurement. By mid-October, the system had already helped trigger 12 antitrust cases. Going forward, the agency expects not only to detect known forms of bid rigging but also to use neural networks to uncover new ways bidders coordinate their actions in tenders.
