DATACROS Through a Journalist’s Lens: The

“I don’t believe anything I read, and only half of what I see.” This sentence – one of my dad’s many quippy, cynical mantras – often floats into my mind while I work. He’d say it most often when we watched the news together. “Figure it out for yourself,” he’d say. It’s perhaps not surprising, then, that I became an investigative reporter. 

As a data journalist for the Organized Crime and Corruption Reporting Project (OCCRP), I’m part of a global network of investigative reporters whose mission to expose wrong-doing at the intersection of money and power. We coordinate and participate cross-border investigations intent on uncovering corruption around the world. Our stories cover anything from drug deals to bid-rigging, political bribes, environmental crime, fraud call centre networks, and more, and our coverage spans jurisdictions around the globe. Our emphasis on cross-border methods has only become more relevant since OCCRP’s inception, as money laundering operations and similar corrupt practices become increasingly international in scope. 

Within this context, myself and the rest of the data journalism team focus on applying the best data-driven methods available to our work. In practice, this can take many forms. One day, we’ll be asked to scrape data off of a website. The next day, we’ll be sent a bunch of unstructured Excel files and asked if we can combine them, clean them, and get a summary statistic out of it for the landing page of an imminent publication. Other days, we’re processing a new batch of leaked data into a navigable format for all the journalists collaborating on a given project. Some common questions the data journalism team gets include: “is there a way to automate this?”, “What can and can’t I say about these numbers, am I reading this right?”, and “what’s the most reliable data source for the topic I’m researching?” Beyond this day-to-day, our team is always on the lookout for software or tooling that will help bolster our findings without compromising the quality of our work. 

DATACROS definitely fits the bill. The platform’s ability to detect anomalies in corporate ownership structures – particularly its focus on legitimate businesses – falls squarely within our remit as investigative reporters. Much of investigative journalism is, in essence, anomaly detection. A lot of this work can also be arduous and repetitive: digging through a suspicious company’s financials hoping someone slipped up, building a network diagram, or mapping cryptocurrency transactions. In some cases, the most time-consuming tasks are the most fruitful: I recently combed through weeks and weeks of inter-office chats in the hopes an employee would accidentally use someone’s real name instead of the alias the organisation had assigned to them (they did). Having the time to do this as a time-strapped journalist can feel like a luxury. So, when a tool comes along that automates even parts of the investigative process, I’ll jump at the chance to use it. 

In its third iteration, the DATACROS tool’s ability to automate these complex processes allow the user to easily visualise corporate links and break down risk factors. The tool is also being enhanced with new AI features, including NLP-based analysis of information extracted from unstructured data sources, satellite imagery processing, and the application of machine learning techniques to process and understand large-scale datasets on assets and financial domains.

On OCCRP’s own data platform, Aleph, we help reporters “follow the money” by granting them access to an array of useful databases by making them searchable. Like DATACROS, Aleph’s purpose is to fill a dearth in tooling for an underserved industry, and is specifically built with investigative journalists in mind. Reporters mine the data on it for stories, and then build out their investigation within Aleph by building network diagrams, timelines, cross-referencing their own uploaded data, and much more. Our work is frequently carried out asynchronously, and, try as we might to follow our project plans, processes often get derailed and a degree of chaos is to be expected. 

At the risk of being slightly reductive, most people who choose to be journalists share a few core character traits: we’re curious, persistent, cynical, and harbour a healthy distrust of authority – or, at the very least, a modicum of discomfort. Narrow the selection down to those of us investigating corruption, and that discomfort can balloon, as can the cynicism. When I first met the DATACROS coordination team at Transcrime-Università Cattolica del Sacro Cuore, the cynicism began to evaporate. The team’s focus, composure and conscientiousness put me immediately at ease. What’s more, being able to speak to a mixture of law enforcement, regulators, lawyers, academics, and private sector workers all at once felt like a breath of fresh air. I left feeling invigorated at the possibility of working alongside everyone while maintaining the independence required of journalists. 

The people who work for journalistic organisations share many traits with public authorities. The staff is frequently comprised of qualified individuals who could be making more money in the private sector, but choose to do the work they do because they’re motivated by the mission or purpose of the organisation. There’s a lot of overlap in the work of journalists and public authorities as well. We both operate on behalf of public interest (“how does this pass the public interest threshold?” is a common question in editorial meetings). Like authority staff, we scour corporate records, leaks, public datasets, and much more to uncover clues of a potentially unreported suspicious links and activities. While we have different modus operandi and are subject to different laws, our motivation remains the same. 

It’s never been an easy time to be an investigative journalist, but this year’s definitely not one of the easy ones. Public trust in news continues to decline, as do the industry’s revenue streams. The cuts to USAID decimated newsrooms across the world, including a complete defunding of many journalists who’d put themselves in danger for years in increasingly hostile environments. Now more than ever, we need all the help we can get.