20-year-old Brazilian creates AI that detects corruption. Understand.
Amid the rapid advancement of artificial intelligence in the corporate world and digital consumption, a Brazilian decided to apply the technology to an area that has historically been sensitive in the country: the oversight of public authorities. The developer Bruno CesarAt just 20 years old, he created an AI-based tool capable of cross-referencing large volumes of public data to map potential financial risks involving politicians.
By automating the cross-referencing of information available on official portals (asset declarations, public contracts, business affiliations, electoral records, and corporate data), the system seeks to identify atypical patterns, inconsistencies, and potential conflicts of interest. It is not an accusatory tool, but an analytical mechanism that generates risk indicators and points out where there may be something that warrants further investigation. Come and understand how this idea came to life and what its potential is.
How AI works
The programmer Bruno Cesar, graduated in Computer Science from UFRN (Federal University of Rio Grande do Norte), developed the tool by combining data science techniques, language models, and statistical correlation systems. The project is based on the integration of multiple public sources: electoral courts, transparency portals, commercial boards, official gazettes, and federal business registration databases.
AI collects this information, organizes the data into a relational structure, and then cross-references elements such as Names, CPF/CNPJ numbers, corporate affiliations, declared asset evolution, and contracts signed with public authorities.Based on these connections, the system builds a kind of "relationship map" that reveals structures that would be difficult to perceive in an isolated manual analysis.
The difference lies in the use of language models To interpret context. It's not just a search for name matches; the system analyzes descriptions, positions, dates, and indirect relationships, assigning different weights to each type of connection. This allows for the generation of a risk score based on multiple combined factors.
According to its creator, the goal was never to replace regulatory bodies or produce automatic judgments. The tool acts as an intelligent filter, capable of reducing analysis time and directing human attention to potentially relevant points.
The following diagram illustrates the simplified operation of the tool developed by Bruno Cesar And how is it able to assist in monitoring the use of public funds?
Recently, Bruno César made the entire tool available at open code in your repository at GitHubThere you can follow the instructions to download and configure the necessary files and test the tool directly on your machine.
Success stories
Although the tool created by Bruno César is still expanding, some practical tests have already demonstrated how the automated cross-referencing of public data can reveal patterns that would normally go unnoticed. Below are some of the most relevant examples.
Case 1: Detection of possible self-directed parliamentary amendments
One of the most impressive examples of how the tool works is its ability to... Mapping complex relationships between parliamentary amendments, public contracts, and companies linked to politicians' family members.In the first case identified by the system, the platform automatically analyzed data from public portals and detected a pattern that may indicate self-direction of public resources.
AI identified that a federal congressman allocated approximately R$ 47 million in parliamentary amendments for a specific municipality.Based on this initial information, the system cross-referenced data from public contracts, business records, and family ties present in open government databases.
The result was the identification of a worrying pattern: Approximately 67% of the contracts linked to the amendments were executed by companies connected to the parliamentarian's own family.Among the links detected were companies associated with the brother and also businesses linked to the politician's son.
The tool not only points out the problem, but also visually explains the chain of relationshipshighlighting the identified pattern:
Parliamentary amendment → Allied city hall → Contracts with family-owned companies
This visualization is presented in a connection graphThis allows journalists or investigators to quickly understand how public money circulates among different actors. Furthermore, the system indicates the sources used in the analysis, including databases such as:
The algorithm also assigns a risk level, which in this case was classified as "critical", with about 97% probability of irregularity, according to the identified patterns.
This type of analysis, which would normally require weeks of manual investigationThis can be done in just a few seconds using this tool.
Case 2: Discovery of networks of administrative irregularities
Another example shown by the tool reveals the system's capability to to identify multiple signs of irregularity within the same term of office.This is something that often goes unnoticed when public data is analyzed separately.
In the panel displayed by the platform, the AI presents a list of alerts classified by risk level, highlighting potential problems detected after cross-referencing public databases.
The system found evidence of the existence of 34 ghost employees, which would represent an estimated cost of approximately R$ 2,4 million per year in public fundsThis type of irregularity occurs when people are registered as employees in government offices or agencies, but They do not actually perform any professional activity..
AI identifies this pattern by analyzing data such as:
- payment records
- employment relationships
- professional activity history
- family and political relations
When this information conflicts or reveals inconsistencies, the system generates an alert.
Another alert identified involves the transfer of approximately R$ 800 for a school that shows strong signs of being non-existent or inactive.The tool detected inconsistencies between educational records, public contracts, and administrative data.
Furthermore, AI also suggests a possible A closed circuit between electoral donations and public funding in the health sector., involving about R$12 million in public transfers and approximately R$150 in political donations..
In this case, the tool identifies when Companies that receive public contracts later appear as donors to political campaigns., creating a potentially irregular funding cycle.
Finally, the platform also pointed out a significant discrepancy between the declared assets and officially registered income, classifying the case as high risk. In the example analyzed, a politician with a declared net worth of R $ million 2,8 It appears associated with financial transactions and assets that may exceed R $ million 80.
This type of analysis is possible because the system cross-references data from multiple sources, such as:
- election statements
- tax records
- associated companies
- public contracts
- political donations
By consolidating this information into a single dashboard, the tool is able to reveal invisible patterns in traditional analyses, making it easier to identify potential corruption schemes.
Practical applications
The tool's potential extends beyond isolated success stories. It can be used as a continuous monitoring and analysis instrument, both by media professionals and civil society organizations.
Corruption oversight
The most obvious application is in fight against corruptionBy automating complex cross-referencing of financial data and business relationships, AI can function as an early warning system. It doesn't accuse, but signals patterns that deviate from expected behavior.
In a country where investigations often face difficulties analyzing large volumes of data, automation can represent a significant gain in efficiency. The use of this tool by regulatory bodies, should it occur, could reduce screening time and expand the scope of oversight.
Support for journalists and civil society
For investigative journalism, technology represents a strategic advancement. The most laborious stage of many reports involves precisely the cross-referencing of data scattered across multiple databases. By automating this phase, AI allows journalists to focus their efforts on contextualization and in-depth reporting.
In addition to identifying suspicious patterns, the tool has also demonstrated efficiency in the automated generation of structured reports. Journalists who tested the system reported a significant reduction in the time spent gathering preliminary data. Civil society organizations can also benefit from the system, using the generated reports as a starting point for social control initiatives and monitoring of public policies.
Transparency in campaigns and mandates
During election periods, cross-referencing campaign donations, business ties, and public contracts can offer voters a broader view of potential conflicts of interest. This tool allows for the analysis not only of isolated declarations, but also of the financial ecosystem surrounding candidacies and mandates. This contextualization capability broadens the debate on transparency and public accountability.
Brazilian Accelerationism
The emergence of tools like AI created by Bruno Cesar It also engages with a broader intellectual movement that has been gaining ground in the international technological debate: so-called accelerationism, especially in its most recent form known as effective acceleration.
O accelerationism It originally emerged as a philosophical movement that advocated for accelerating technological and economic processes to bring about profound social transformations. According to the WikipediaThe concept gained prominence in the 1990s and 2000s, inspired by thinkers such as the British philosopher... Nick Land, who argued that technological advancement and the growth of capitalism could lead to radical changes in the structure of society.
Also frequently abbreviated as e/accThis movement emerged in the 2020s within online technology communities and advocates for an explicitly techno-positivist stance, encouraging the rapid advancement of technology, especially AI, as a way to solve global problems such as poverty, energy crises, and environmental challenges.
The defenders of e/acc They believe that accelerated technological progress is inevitable and desirable, and that attempting to curb this process through excessive regulations or fear of technology could delay important solutions for humanity. In many cases, this view connects with ideas of transhumanism and advanced technological civilizations capable of expanding human knowledge and utilizing more energy and resources on a global scale.
When talking about Brazilian accelerationismHowever, the concept takes on a more pragmatic interpretation. Instead of discussing futuristic scenarios about artificial general intelligence or interplanetary civilizations, the idea appears linked to the intensive use of technology to solve the country's structural problems — such as corruption, public transparency, and access to information.
In this context, tools such as AI developed by Bruno Cesar They represent a kind of civic accelerationism: the use of algorithms and automation to expand society's investigative capacity. By cross-referencing large-scale public databases, these technologies allow journalists, researchers, and citizens to identify patterns that previously would have required months of manual work.
In other words, it's about using the power of technology to accelerate democratic oversight, making the monitoring of public power more efficient and accessible. While in Silicon Valley accelerationism is often associated with advances in artificial intelligence and the digital economy, in Brazil it can take on a different role: serving as a tool to strengthen transparency, accountability, and citizen participation.
The tool created by Bruno Cesar This shows that artificial intelligence can go beyond corporate automation or text generation. When applied to the analysis of public data, it has the potential to strengthen transparency and expand the capacity for social oversight.
However, like any high-impact technology, it demands responsibility, human validation, and ongoing ethical debate. The challenge now is not only technical but also institutional: how to integrate such tools constructively into the democratic ecosystem?
So, do you think this tool can really help in the more efficient fight against corruption? Leave your opinion in the comments.
Learn more
Sources: Examination, Tecmundo
reviewed by Tiago Rodrigues in 05 / 03 / 2026
Discover more about Showmetech
Sign up to receive our latest news via email.
