evidoria

← Back to browse

Good practice Imported

BRAVA — Spain's Competition Authority Uses AI to Catch Bid-Rigging Cartels, With Sanctions to Prove It

Spain · Madrid · See the Spain profile · See the Madrid profile

Evidence: Descriptive / self-reported Top 46% 60/100 · Ask Evidence Copilot about this practice

CNMC's BRAVA tool has used machine learning to flag collusive bids in Spanish public tenders since 2024; President Cani Fernández says BRAVA-flagged cases have led to sanctions, and CNMC is now training the US DOJ and building a successor tool, ATENEA.

BRAVA — Spain's Competition Authority Uses AI to Catch Bid-Rigging Cartels, With Sanctions to Prove It

Details

Maturity
Scaling
Promoter
Comisión Nacional de los Mercados y la Competencia (CNMC), Economic Intelligence Unit
Period
2024–2026
Keywords
competition & antitrust enforcement, public procurement integrity

Context

Spain's national competition authority, the Comisión Nacional de los Mercados y la Competencia (CNMC), built an in-house machine-learning tool called BRAVA (Bid Rigging Algorithm for Vigilance in Antitrust) inside its Economic Intelligence Unit, screening public-tender bids for signs of collusion from 2024.

Activities

BRAVA classifies submitted bids as potentially collusive or competitive, automating a task that previously depended almost entirely on manual document review or a prior tip-off.

Results

In February 2026, CNMC president Cani Fernández stated that BRAVA-flagged leads have already resulted in sanctions, citing a resale-price-maintenance sanction against I.C.O.N. and a bid-rigging cartel in the travel-services sector; she also said CNMC is training the US Department of Justice on the method and building a follow-on reasoning tool, ATENEA.

Conclusions

As with comparable tools piloted at Lithuania's Competition Council and Brazil's CADE, CNMC has not published BRAVA's precision or false-positive rate, nor how many sanctioned cases were BRAVA-originated versus flagged through other channels — the public record so far is the regulator's own account.

Implementation
Indicative cost
Medium (€50k–€500k)
Time to results
Medium (1–3 years)
Staffing & skills
CNMC Economic Intelligence Unit

Conditions for success

  • In-house machine-learning capability inside the regulator
  • Access to public-tender bid data

Common failure modes

  • No published precision or false-positive rate; cannot isolate BRAVA-originated sanctions from other detection channels

Commonly funded by

Digital Europe Programme

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

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