Anonymizing Data by the Book: From Ambiguous to Clear Data Anonymization
These days, many organizations are grappling with a question that, surprisingly, is quite complex: What data are you still allowed to retain? and which must be deleted? For example, housing cooperatives, educational institutions, healthcare organizations as well as HR departments, this is not a theoretical discussion. It concerns specific cases, personal, privacy-sensitive data and processes that have been in use for years. But is it permissible for what youyou and your employees do still allowed?
On Unclear Data Anonymization
Processes often need to be re-evaluated, particularly due to increased regulatory pressure, such as the GDPR, NIS2, and sector-specific guidelines like BIO or BBV/BSV. If you retain data that you’re not supposed to, it can lead to hefty fines or tough conversations with auditors. Even worse can be the loss of customer trust, which results in a loss of revenue. But how do you ensure that you’re complying with the rules?
Why Make Decisions Manually?
Sometimes it seems simple to let employees decide for themselves what can and cannot be retained. In practice, this rarely works. For example, an employee who opens an old tenant file might see a document containing Social Security numbers. Is it necessary to detect fraud, or can this information only be used when registering for a rental property? Or consider an HR employee who comes across a payroll file from five years ago: can it be retained for analysis, or must it be deleted? It’s unrealistic to expect an employee to be familiar with every letter of the law—that would require everyone to be a lawyer or have a genius-level intellect.
And then there are the outdated files…
Records from years ago pose an even greater challenge. Often, these records were not organized in accordance with current regulations. This frequently leads to frustration and internal disputes. Sometimes it is unclear where the source is located and which data may still be retained. In education, for example, an old student ID might remain on file simply because no one realized it wasn’t allowed. HR departments sometimes discover contracts or medical records that, legally speaking, should have been deleted long ago. Data rarely stands alone, which makes it even more challenging. It is interwoven into reports, analyses, and processes. This creates risks without anyone even realizing it. Are you absolutely certain that no privacy-sensitive information is stored somewhere where it shouldn’t be?
The solution does not begin with assessing documents on an individual basis, but rather with examining the entire process of incoming and outgoing information flows. A baseline assessment provides insight: where is the data located, what rules apply, and how does information flow through the organization? With that insight, you can determine what can be securely anonymized, what needs to be deleted, and how processes can be designed to mitigate future risks. It’s extremely helpful to have a specialist review the situation. This is someone who understands both the processes and the legislation. This way, working in compliance isn’t a gamble, but a structured process centered on clarity and control.
Automatic anonymization
At OneXillium, data anonymization begins the moment information is received. Using the Digital Orchestration Platform (DOP), documents are automatically recognized, classified, and routed through the appropriate workflows, with data anonymization applied immediately. Manual decisions and checks are no longer necessary, while the organization remains fully compliant. As a result, anonymization is no longer a standalone task or a source of uncertainty, but an integrated and reliable process that works from the very first step.