Can you begin by describing your role as a detective?

Essentially, I serve as an investigator. Brazil’s legal system is intricate, particularly in criminal justice, with segregated institutions and overlapping functions within the same job titles. In summary, I work as an investigator in a department akin to a “detective bureau” in the United States.

What are some of the most complex aspects of investigating drug crimes?

The primary obstacles to drug-related investigations in Brazil stem from the criminal justice system, general criminal procedural laws, and specific drug legislation, all of which are subject to frequent interpretative shifts. Consequently, in Brazil, an investigator might achieve success in an investigation today, only to have all their efforts rendered invalid tomorrow, even if they follow the same procedures as before.

Apart from that, the depth of the investigation for each case is crucial. What I mean is, if a minor drug dealer leaves evidence that leads back to their supplier, the investigation can become intricate, requiring telematics analysis, banking data, surveillance, personal interactions, and other factors. In such scenarios, cooperation from the criminal justice system is vital, particularly for addressing police requests submitted to the judiciary. Delays in this process could hinder timely police investigations.

If everything “goes smoothly,” the primary challenge in a complex case often lies in weaving together ideas, visually illustrating relationships between entities and objects, and securely storing information to uphold the integrity required for maintaining the chain of custody. In essence, the investigators must manage information and evidence carefully to ensure they remain accessible and not compromised by any legal procedural oversight.

How does data play a role in resolving these cases? Can you give an example of a case where data was crucial?

The term “data” is quite broad, so I’ll address two commonly understood meanings in Brazil: a) intelligence data and b) data as informative elements. Intelligence data refers to information collected and generated to support decision-making, which is not necessarily aimed at producing “evidence” for criminal proceedings. In contrast, data as “informative elements” includes information that, after undergoing the “due legal process,” may form part of the evidence list that supports a conviction.

During an investigation, an investigator often encounters both types of data, which, generally speaking, cannot be formally combined, as doing so could invalidate their work. Therefore, effective data management is essential for successful case resolution. Intelligence data typically guides police investigations (even though it cannot serve as evidence), while “informative elements” hold some probative value.

Proper management of this data is critical. For instance, when police officers realize that a specific source of information is of an intelligence nature, they become fully aware of the consequences of its improper use.

What are the main data sources you rely on during an investigation?

Data gathered from closed sources is undeniably valuable. When consolidated, effective data scraping from these closed sources can unveil significant insights about the target.

How do you handle data from different sources such as informants, surveillance and digital records?

As mentioned in earlier responses, the management of information from sources depends largely on its legal classification. Intelligence data is usually generated selectively for investigative reasons. Essentially, managing these sources depends on the expertise of each police officer.

In standard investigations, if the sources are classified as intelligence information, they are likely stored in a personal access file or retained “in the officer’s memory.” Conversely, probative information is formally documented in reports that guide and support the entire police investigation and the subsequent prosecution in court.

In Brazil, law enforcement still struggles to leverage informally collected data that cannot be directly utilized in formal criminal cases. There tends to be a strong focus on procedures that can be formally justified, often resulting in the discarding of significant intelligence information after its initial use.

Regarding surveillance and digital records, Brazil employs a “chain of custody,” which mandates adherence to various legal requirements to ensure the evidence remains valid. In summary, most investigators in Brazil organize their cases using traditional digital storage methods, creating folders and subfolders in operational systems where all types of files are accessible.

How does outdated or incorrect data impact an investigation?

Such errors can result in the wrongful arrest of an innocent individual or the execution of a search warrant at an incorrect location. This may lead to faulty conclusions and inferences, or at the very least, to ineffective or unnecessary actions that hinder investigations. In solving a crime, the longer the duration of the investigation, the more challenging it becomes to clarify the details.

Can you talk about legal restrictions on data use? How do privacy laws affect investigations?

As I noted before, the most significant issue in Brazil is the frequent changes in law interpretation. For instance, consider a scenario where a company has specific information that could assist law enforcement without needing judicial authorization today. However, tomorrow, an interpretation shift might dictate that acquiring this data requires a court order to avoid liability for improper disclosure. A week later, the judiciary might revert to the original interpretation as the correct one.

As a result, companies with sensitive information often overly restrict access to law enforcement as a precaution against potential legal consequences, which hampers investigative efforts. In Brazil, privacy laws do not apply to police forces in certain contexts, particularly where a court order is necessary for data release. The primary hurdle lies in the bureaucracy and the duration between requesting the court order and its execution.

Non-compliance renders data acquisition illegal, resulting in any subsequent evidence derived from it being considered invalid. The core challenge arises with the potential for reusing this information. When judicial authorization exists for accessing protected data, this authorization typically pertains strictly to the specific case. Consequently, if I were to input data into a law enforcement database, utilizing that information formally would require a new court order; otherwise, using it “informally” could jeopardize any future investigation’s validity.

What barriers exist in sharing data between law enforcement agencies in Brazil?

In addition to the legal barrier, as in the example I mentioned above, information systems lack standardization, employing different data storage formats, programming languages, and so on. Furthermore, there is a reluctance to share certain “critical” data, fueled by distrust among different police departments.

How do you manage the volume of data generated in drug investigations?

I separate what I would consider elements of intelligence information into text files, something as a “homemade data lake” for later, individual access. I then organize usable data into folders based on thematic relevance and/or target identification.

What kind of training do detectives receive to work with data effectively?

In Brazil, police officers lack effective training for data management, especially from a managerial perspective, and there is no established methodology.

While some positions require basic knowledge of archival science, these concepts are typically introduced during police academy training rather than taught in a practical, focused manner. Consequently, the ability to manage information largely relies on each officer’s individual skills.

However, there have been advancements in specialized units dealing with areas like cybercrime, embezzlement, financial crimes, and organized crime, particularly through the use of data analysis tools.

What improvements would you like to see in the way data is managed and used in drug investigations?

The establishment of a database that allows for easy information entry with minimal bureaucracy or structured data requirements, similar to a data lake, encompassing intelligence data that can be integrated with reports and other insights gathered from closed sources.

How improved data sets (such as gang profile forms, trafficking routes, transaction hotspots, etc.) could help them connect the dots more effectively and track their organized crime activities in your region?

In my federative unit and police agency, the central intelligence body is the sole entity responsible for comprehensive analysis and information cross-referencing. This is primarily due to the absence of structured information systems that facilitate access for all police officers to vital intelligence data. Addressing this issue could significantly enhance both investigative efforts and the planning of public security policies.

How do you see the future of data-driven drug investigations evolving in Brazil?

Investigations will rely on digital technologies not just for drug-related cases, but across all topics, often referred to in Brazil as “cybercrimes.” These techniques are currently employed in various types of crime, and drug offenses are no exception, especially as the overarching aim is always to generate profit. Nowadays, traces of money without digital footprints—whether through cryptocurrencies or traditional banking—are nearly nonexistent. Cash (paper money) is infrequently used for large-scale criminal activities.

What advice would you give to other law enforcement agencies facing data challenges?

The extensive use of computerized systems for data management, particularly for information that may seem insignificant today, can yield valuable insights when combined with other datasets in the future. These insights can significantly enhance decision-making processes and operational efficiency. To this end, I recommend developing a data lake equipped with advanced data retrieval capabilities powered by artificial intelligence (AI). This system could not only store vast amounts of data but also suggest relevant, pre-existing information tailored to the specific needs of each case, as identified by an investigator. Additionally, I would propose incorporating features for data concatenation, analysis, and interconnected visual representation, enabling a comprehensive and intuitive understanding of the data. This approach would facilitate more informed and strategic decision-making.