Brazil will use artificial intelligence to collect mining royalties

Brazil will use artificial intelligence to collect mining royalties

Brazil’s mining regulatory agency, ANM, will use artificial intelligence to increase efficiency in the task of collecting mining royalties.

The regulator carried out a selection process for initiatives to support improvements in deadline control for mining royalties, known as Financial Compensation for the Exploitation of Mineral Resources (CFEM).

According to the ANM, there is currently a backlog of more than 11 thousand litigation cases and a potential revenue loss of 20 billion reais (US$3.9 billion), considering that there is a deadline for the regulator to collect unpaid royalties.

“The challenge is to strengthen ANM’s capacity to protect public credits and help ensure that the resources from CFEM are collected and distributed to states and municipalities, where they can support public policies and offset the impacts resulting from mining activity,” said Alexandre Rodrigues, ANM’s superintendent of revenue collection and oversight, in a statement.

The ANM initiative to select artificial intelligence solutions is called the Vênia program and its main focus is to support the prioritization of administrative litigation processes of the CFEM.

In the first stage of the program, held between June and July, the Vênia team took part in workshops and practical activities on artificial intelligence, ethics, user experience, and product development. The work resulted in the definition of functionalities, requirements, and risk criteria, as well as the creation and refinement of the prototype.

"The team tested hypotheses, refined the flow, and defined what artificial intelligence can safely do, as well as what must necessarily remain under human analysis," explains Giovanna Brasil Vieira, who is responsible for Vênia’s governance and legal compliance.

Vênia must process documents and metadata, identify relevant dates and milestones, recognize the procedural stage, generate alerts, and prioritize cases according to the risk of missing deadlines.

"In practice, the solution should help the civil servant to locate critical cases more quickly and view the evidence that supports each indication made by the artificial intelligence", explains Edson Matsumoto, the project’s data analyst.

Among the pilot’s goals are achieving at least 85% accuracy in identifying procedural milestones, reducing screening time by up to 90%, increasing analysis capacity by at least 20%, and proactively identifying 100% of the sample cases with critical deadlines.

The next steps include providing a secure environment for the pilot, enabling controlled access to SEI and, when necessary, to other Litigation databases, as well as selecting the sample of cases that will be used in the homologation.

The solution will be launched at Innovation Week 2026, in November.

(The original version of this content was written in Portuguese)

Subscribe to the leading business intelligence platform in Latin America with different tools for Providers, Contractors, Operators, Government, Legal, Financial and Insurance industries.

With the new structure, companies of all sizes now have a more automated operation, enhanced protection for hybrid and multicloud environments, and...

Follow the main articles most recently published by BNamericas.

Subscribe to Latin America’s most trusted business intelligence platform.

Other projects in: ICT (Brazil)

Get critical information about thousands of ICT projects in Latin America: what stages they're in, capex, related companies, contacts and more. Request Demo

Other companies in: ICT (Brazil)

Get critical information about thousands of ICT companies in Latin America: their projects, contacts, shareholders, related news and more. Request Demo