evidoria

← Back to browse

Good practice Imported

BRIAS — South Korea's AI-Powered Bid-Rigging Indicator Analysis System for Public Procurement

South Korea · Sejong · See the South Korea profile · See the Sejong profile

Evidence: Observational / pre–post Top 14% 80/100 · Ask Evidence Copilot about this practice

Since 2006, Korea's Fair Trade Commission has used an automated indicator-scoring system to flag collusive bidding patterns in public tenders. It now screens roughly 60,000 procurement cases a year from over 720 public entities, with fines reported at nearly 40 times its running

332 agencies
Participating public procurement agencies (2014) (2014)
720+ entities
Public entities feeding data into BRIAS (current)
~60,000 cases/year
Procurement cases screened per year (current)
~40x ratio
Fines recovered relative to system running cost (over a seven-year period)
BRIAS — South Korea's AI-Powered Bid-Rigging Indicator Analysis System for Public Procurement

Details

Maturity
Established
Promoter
Korea Fair Trade Commission (KFTC)
Period
2006–present
Keywords
competition enforcement, public procurement, fraud detection, fair trade

Context

South Korea's Fair Trade Commission (KFTC) developed the Bid Rigging Indicator Analysis System (BRIAS) in 2006 to help identify collusive bidding in public procurement, applying a weighted-indicator model to tender data such as bidding method, number of bidders, number of failed bids, and how closely winning bids cluster around the estimated price.

Objectives

BRIAS scores each tender's likelihood of bid rigging so that KFTC investigators can prioritise scarce investigative resources on the cases most likely to involve collusion.

Activities

BRIAS was first applied to South Korea's central purchasing body in 2006, then extended in 2007 to four major state-owned enterprises. Flagged tenders are referred to KFTC investigators for further review; BRIAS does not itself determine wrongdoing.

Results

By 2014, 332 public procurement agencies were participating, according to the OECD's review of the programme. More recent reporting by the Open Contracting Partnership puts BRIAS's data feed at over 720 public entities, screening roughly 60,000 procurement cases a year, with KFTC investigators reporting fines nearly 40 times higher than the system's maintenance cost over a seven-year period.

Conclusions

The OECD has cited the programme as a model for other competition authorities seeking replicable, rules-based tools to prioritise scarce investigative resources; because scoring weights and case-level outputs are not made public, external verification of individual flags is limited to what KFTC and the OECD publish in aggregate.

Implementation
Indicative cost
Low (< €50k) — KFTC reports fines recovered were nearly 40 times the system's running cost over a seven-year period, though the absolute maintenance cost is not published.
Time to results
Long (> 3 years) — First applied in 2006 to Korea's central purchasing body, extended to major state-owned enterprises in 2007, reaching 332 agencies by 2014 and over 720 entities today.
Staffing & skills
Korea Fair Trade Commission (KFTC) investigators who review and act on flagged tenders

Conditions for success

  • Human referral and review of every algorithmically flagged tender before enforcement action is taken
  • Consistent access to tender and bid administrative data across public procurement
  • Sustained institutional ownership by a statutory competition authority over multiple decades

Common failure modes

  • Scoring weights and case-level outputs are not made public, limiting external verification of individual flags.

Where it fits

Governance type
national statutory competition authority
Scale
national
Income level
high-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Replication kit

Reusable artefacts from this practice — as published by their sources.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful