MMOOZZGG.com

AI against corruption

South Korea bid rigging: how BRIAS has searched for it since 2006

5 min read2 sources

BRIAS is the system South Korea uses to find bid rigging in public tenders by bid statistics: who bids at the same time, who always offers a low price, and how prices change over time. The Korea Fair Trade Commission (KFTC) runs it, and since December 2023 more than 700 organizations must send tender data to it.

On this page · 6 sections

In short

BRIAS is a system of the Korea Fair Trade Commission (KFTC) that has collected public tender data since 2006 and searched it for signs of bid rigging in South Korea. In 2023 it ran 40,170 analyses, against 25,520 in 2018. BRIAS does not prove collusion: it gives investigators leads, and people draw the conclusion.

How BRIAS works

The KFTC explains why Korea needs a separate system with an OECD figure: removing bid rigging cuts procurement costs by 20% or more. Collusion raises prices for the state, wastes the budget and undermines public trust.

BRIAS (Bid-Rigging Indicator Analysis System) has worked since 2006. It is a database that brings together information on tenders of public organizations and computes statistics from it.

Not all contracts are analyzed, only large ones: from 100 million won for goods and services, from 5 billion won for general construction and from 500 million won for specialized construction.

Buyers send tender data to the KFTC, and the commission sees it in one place. For the KFTC, it is one of three tools against collusion, together with a leniency program for a participant who reports a cartel and rewards for informants.

Coverage was widened by an amendment to the competition law in December 2023. The organizations that send data now include local public corporations and quasi-government institutions, more than 700 in all. The KFTC writes that this closed blind spots in procurement monitoring.

Which signs it counts

The KFTC names three patterns that BRIAS helps to see in bids:

1. Different participants submit bids at the same time.

2. One company always offers a low price, and the others are clearly higher.

3. Bid prices move unusually over time.

Each sign on its own can have an explanation: companies may bid on the last day of the deadline, and prices may rise because of raw materials. A sign becomes a signal for a check when it repeats across many tenders with the same participants.

The KFTC compares BRIAS to surveillance cameras: it is a monitoring tool, not instant punishment, and the camera covers every street, including narrow alleys. The full set of bid rigging signs is covered on the page about bid rigging.

Growth from 25,520 to 40,170 analyses

The number of BRIAS analyses grows every year: 25,520 in 2018, 28,494 in 2019, 29,343 in 2020, 30,496 in 2021, 30,893 in 2022 and 40,170 in 2023.

In its presentation, the KFTC calls this figure the number of analyses. It does not say how many of them refer to separate tender procedures, so it is more accurate to speak of analyses rather than tenders.

The sharp increase came in 2023. In December of that year, the group of organizations required to send data grew to more than 700, so the system’s coverage became wider.

People cannot handle such a flow: 40,170 analyses a year is more than 100 a day. The machine computes statistics for each tender, and an investigator looks only where the pattern resembles collusion.

An example case

In its presentation, the KFTC describes a case about purchases of inorganic coagulants. These are chemicals used at water treatment plants to separate solid particles.

Several water treatment corporations ran a series of purchases in 2019–2023. After an investigation, the KFTC found that two companies had agreed in advance to split volumes and set prices: one won 14 contracts and the other 17, worth 438 billion won in total, or 43% of the market.

The KFTC used BRIAS as a database along the whole path of the case: from the first suspicions during monitoring to the final confirmation of collusion. The outcome: orders and a fine of 391 million won, about 273,000 dollars.

Note the order. The system gave data and statistics, and the KFTC investigation, with its powers and evidence, confirmed the collusion.

Why the system gives leads, not proof

The KFTC says it directly: BRIAS cannot prove an agreement, but it gives investigators important leads. Matching prices or bid times are a sign of risk, not proof of collusion.

Proof requires what tender data does not contain: correspondence, testimony, documents about the agreement. An investigation obtains these, the KFTC decides on a violation, and a court settles any dispute about it.

If collusion is proven, the KFTC can issue orders, fines and sanctions, such as limiting a company’s participation in tenders. Such consequences require evidence, not a statistical match.

The system also has a deterrent effect. The KFTC notes that BRIAS shows business that tenders are under constant watch, and this is a strong argument against collusion.

In addition, BRIAS data and statistics support informants’ complaints. A person reports a suspicion, and the commission sees at once whether the tender figures back up the claim.

What Russia can take from it

In Russia, the Federal Antimonopoly Service (FAS) handles a similar task. The first module of its Big Digital Cat system started in autumn 2019, and since 1 August 2025 the Anti-Cartel state information system has analyzed data from the Unified Procurement Information System and the Torgi system every day with elements of machine learning, covering 100% of procurement procedures, according to FAS.

The Korean experience adds four decisions to this:

1. An entry threshold. BRIAS analyzes tenders above set amounts, and investigators focus on large contracts.

2. A duty to send data. In Korea, the law widened the group of buyers so that no blind spots remain.

3. Links to other tools. System signals work together with the leniency program and informants’ complaints, not instead of them.

4. Work with buyers. The KFTC meets procurement bodies regularly, helped them adopt internal rules against their staff taking part in collusion and supported their own anonymous reporting lines.

How control systems work in Russia and the CIS countries is covered on a separate page, and you can compare Korea with other countries on the page about global experience.

Sources and fact-checking

Checked by the editors on 16 October 2026.

  1. KFTC, presentation for the UNCTAD working group on cross-border cartels, 10 April 2025 (unctad.org)
  2. Interfax, 31 May 2021; Rossiyskaya Gazeta, 14 July 2025; ComNews, 22 December 2025 (comnews.ru)

Found a mistake? Tell us

Publications

Mike Mozg

Agentic AI use cases at AT&T: nearly 500 processes now run on agents

AT&T no longer uses AI only to help staff work faster: it is rebuilding whole business processes around agents. SiliconANGLE reported on 14 November 2026 that nearly 500 agentic AI use cases at AT&T are already in production. Here is how the company rebuilds them, where a human still checks the agent, and what other large companies can take from it.

News4 min

For government bodies

We build such systems for government bodies and state companies, from data analysis to a pilot on one scheme. The data stays on the organization's own servers.

Contact us