Background: an urgent but familiar problem
The problem is well documented: criminal investigations are experiencing significant delays because filtering legally privileged information takes too long. In 2024, the Dutch Supreme Court (Hoge Raad) signalled that it is for the legislature to act. Later that year, the Research and Documentation Centre (WODC) published its comparative study on professional legal privilege, intended to provide inspiration and practical guidance for improving the treatment of legal privilege in criminal practice. That report was one of the direct catalysts for the Bill now under consultation.
Before turning to the proposed reform, we first consider the current framework and the legitimate criticisms levelled against it.
Current practice: the examining judge in charge
Under the current system, the examining judge has overall responsibility for the filtering of privileged material. The actual filtering is carried out, largely through automated processes, by law enforcement officers who have been sworn in by the examining judge as ‘privilege officers’ (geheimhoudersfunctionarissen). Crucially, these are officers with no substantive role in the investigation at hand, operating under the authority of the examining judge.
This arrangement has been criticised for years, and rightly so. In 2022, the Netherlands Bar Association (NOvA) objected to the involvement of law enforcement officers in filtering privileged information. A privilege officer performs this role alongside substantive work on other investigations, creating an inherent conflict of roles. The WODC confirms that the position is organisationally complex: privilege officers belong to the investigative apparatus but are functionally accountable to the examining judge. That tension makes the role unattractive and leads to capacity shortages, particularly in smaller investigative agencies.
We fully endorse this criticism. That reform is needed, is beyond question.
The proposal: filtering shifts to the public prosecutor
The question, then, is what form that reform should take. The Bill introduces a two-phase model. In the first phase, the public prosecutor is responsible for a preliminary selection of so-called ‘bulk data’, large, unsorted datasets, with access to content kept to a minimum and the selection process automated wherever possible (Article 2.7.61b). In the second phase, the assessment procedure, the examining judge enters the picture. This happens only where there is a concrete and particularised reasonable suspicion that specific data qualifies as privileged (Article 2.7.61a(2)). The examining judge is thus, in principle, effectively sidelined from the filtering process.
Whether the filtering is carried out by officers who are not involved in the investigation, as is currently the case, is left open by the Bill. That question is delegated to a future Order in Council (Algemene Maatregel van Bestuur).
The Bill disregards the WODC’s recommendations
The legislature has set aside the WODC’s findings and recommendations without substantiation. The WODC advocates strengthening judicial capacity, fostering greater cooperation between the prosecution service and the Bar on filtering practices, and assigning a more prominent role to the examining judge and the court’s advisory chamber (raadkamer) when the parties cannot reach agreement. Where an initial sift by someone other than the examining judge is necessary, the researchers look to the privilege holder or a representative of the relevant profession, not to the investigative authorities involved in the case.
The legislature has nonetheless opted for a radical solution that directly contradicts these recommendations, justifying its choice solely on the ground that the examining judge must be relieved of the burden. The Explanatory Memorandum contains no assessment whatsoever of the suspect’s interest in an independent filtering procedure, nor any reflection on whether the position of the defence is unnecessarily weakened by entrusting the filtering to the prosecution. That omission is striking.
It is, in our view, a proposal that is difficult to justify on any level and that leads to an unacceptable erosion of legal privilege.
Objection I: legal privilege must be surrendered unnecessarily
Automated filtering requires the public prosecutor to know which search terms to apply. In practice, this means the suspect or the defence must disclose which lawyers the suspect has consulted, over what period, and through which channels. But that very information is itself protected by legal privilege. Compelling its disclosure is therefore already at odds with the right to privilege.
Objection II: automated filtering is structurally inadequate
The Bill proceeds on the assumption that automated filtering will result in privileged information being ‘almost entirely filtered out without its contents being examined’. This reflects an overly optimistic and simplistic view of practice.
In financial and economic criminal cases, it is the rule rather than the exception that vast quantities of digital data are seized, often running from hundreds of gigabytes to several terabytes. Automated filtering typically relies on email domain extensions or the names of privilege holders. That works reasonably well for email correspondence with a lawyer. But legal privilege also extends to internal memoranda, minutes and other documents in which legal advice is embedded or which were prepared for the purpose of obtaining such advice. The privileged nature of those documents is not necessarily discernible from their form or file format, and they are not automatically caught by keyword-based filtering. Effective filtering in such cases requires more refined search terms and substantive review of the data.
The Bill allows for substantive review by law enforcement. It introduces a ‘general duty of care’ for investigating officers (Article 2.7.57a), requiring them to exercise due caution when reviewing data. But a duty of care is vague, difficult to enforce, and its practical application is, from the defence’s perspective, entirely unverifiable. Practice has repeatedly shown that investigating teams do access privileged information and use it as a starting point for their inquiries.
Objection III: no safeguards whatsoever for the defence
The current model affords the defence some degree of oversight over the filtering carried out under the examining judge’s supervision. The Bill is silent on this point. It provides no procedural safeguard enabling the defence to verify whether the prosecution’s filtering was carried out correctly and completely. In our view, the defence is thereby forced to rely on the diligence of the very party that has an interest in being able to use as much information as possible. That is not the hallmark of a system governed by the rule of law. We will explore this issue further in a forthcoming blog.
Objection IV: a cosmetic fix for the delay problem
The central argument for the Bill is a practical one: the examining judge must be relieved of the burden of filtering privileged information. But this argument overlooks a critical fact: under the current model, the filtering is already largely carried out by law enforcement, through the privilege officers. That has not shortened processing times. We see no reason to assume that the same work, performed by the same type of officials but now under the responsibility of the public prosecutor rather than the examining judge, will somehow produce faster results. The prosecution service, too, faces chronic high workloads and capacity constraints. Shifting the filtering process to the prosecution displaces the capacity problem; it does not solve it.
Conclusion: invest in the examining judge
Driven primarily by practical considerations, the legislature has put forward a proposal that fundamentally undermines the rule-of-law protection of legal privilege and will needlessly harm the position of the suspect. It is our firm conviction that the independent role of the examining judge in filtering privileged information is of fundamental importance to the right to a fair trial. The Bill disrupts that balance: the defence becomes dependent on the prosecution and law enforcement, parties whose primary concern is the investigative interest, for a process that is inherently unverifiable and that compels the disclosure of information that ought to remain protected.
The solution to lengthy processing times does not lie in shifting responsibility to the prosecution, but, in line with the WODC’s recommendations, in structural investment in judicial capacity, in adequate technical support, specialised software, and the designation of a specialist examining judge’s office for privilege-related matters. That requires resources, but it is the only path that does justice to both the efficiency of the criminal process and the rights of the defence.
After all, no one has ever benefited from the fox guarding the henhouse.
Want to know more?
In our blogs, we discuss in turn the proposal to assign the filtering procedure to the public prosecutor, the proposal to introduce a retention obligation for privileged information found by the Public Prosecution Service and investigating officers, and the lack of oversight mechanisms for the defence over the public prosecutor’s filtering process.
Do you have questions about the implications of this proposed legislation? Our Corporate Crime & Business Integrity team is happy to advise.
