Abstract
Academic literature in England and Wales and New Zealand does not consider the protection of legal professional privilege where digital material is seized under a search warrant. Academic literature in the United States does engage with this subject but is not informed by a comparative approach. This article fills both gaps. It examines practices that have been developed by investigative teams and prosecuting authorities in all three comparator jurisdictions in their attempts to provide safeguards necessary to preserve privilege. Such practices involve the use of technology to increase the speed, cost effectiveness and/or efficiency of identifying privileged documents. The process of developing these practices has been informed by judicial guidance, where they have been challenged before the courts, and by guidance from government departments, Bar Associations or Law Commissions. Following comparative analysis, the article recommends measures that should be included in legislation, codes of practice or guidance in any jurisdictions where there is potential for legal professional privilege or an equivalent concept to be undermined when digital material is seized under a search warrant.
Keywords
Introduction
The principles governing the operation of legal professional privilege date back to an era when communications were made or documented in handwriting. More recently, disclosure of documents in criminal proceedings related to the time-consuming examination of files of typed paper documents. In the digital age vast numbers of documents are now both created and saved in digital form. This increases the potential complexity and cost of disclosure. It also has ramifications for the protection of privileged communications. In the context of search warrants, huge amounts of digital material 1 may be seized with the danger that privileged material may fall into the hands of investigative teams and prosecuting authorities. The storage of large quantities of material on devices such as phones, tablets, hard drives, on-site servers or remote servers is inherently likely to lead to the over-seizing of material, with consequences for the protection of privileged documents during the sifting process. Legislation concerning the scope of search warrants makes provision for excluding privileged material from their ambit. It is crucial that practice both around applications for search warrants (including their drafting and approval) and concerning the use of technology to search and sift privileged digital material keeps pace with technological developments relating to the storage and communication of such material.
Statutes authorising the issuance of search warrants in England and Wales, New Zealand and the United States of America (the United States) do not lay down requirements concerning matters such as the use of keyword searches to filter privileged material. Practices that have been developed by investigative teams and prosecuting authorities are relied upon to provide the safeguards necessary to preserve privilege where search warrants authorise the search and seizure of digital material. The development of these practices is informed by judicial guidance where their implementation, or attempts to implement them, have resulted in challenges before the courts. It is also informed by guidance provided by government departments (e.g., the Department of Justice [DOJ] in the United States), Bar Associations (e.g., the Bar Council of England and Wales) or Law Commissions (English and Welsh and New Zealand). Such practices increasingly involve the use of technology itself to increase the speed, cost effectiveness and/or efficiency of identifying privileged documents and safeguarding privilege therein.
In this novel and significant article, practices adopted by specific investigative teams and prosecuting authorities in England and Wales, New Zealand and the United States are compared as exemplars, alongside guidance provided by the courts, Law Commissions and the DOJ. From this comparative analysis recommendations are made which, if followed, will reduce the risk that privileged material will fall into the wrong hands following execution of a search warrant. The analysis takes into account current limitations of keyword searching and predictive modelling, the necessity of human involvement in the search process and the importance of using personnel to filter material for privilege who are independent of the investigative or prosecuting authority. It also recognises that any processes which are adopted must be realistic and proportionate in terms of time and cost. The topography in this sphere is particularly complex because regulations and practice frequently lag behind technological developments in the storage of data in digital form.
The subject matter of this article is not dealt with by legal academic literature in England and Wales or New Zealand and judicial guidance is extremely limited in those jurisdictions, although both have attempted to grapple with the challenges through Law Commission consultations and reports. Whilst literature and judicial guidance are more extensive in the United States, these have not been informed by comparative methodology. This article fills crucial gaps in the existing academic literature in all three jurisdictions by dealing with a subject that legal academics in England and Wales and New Zealand have not tackled and by providing valuable comparative insights which the academic literature in the United States lacks. The recommendations made in this article will influence knowledge and practice not only in the three jurisdictions considered but also in any other jurisdictions where there is potential for legal professional privilege or an equivalent concept to be infringed in the context of search and seizure of digital material.
The article's focus is on powers of search and seizure under warrant. There are, of course, instances where search and seizure may take place outside the warrant process. These are outside the scope of this piece. Also outside its scope are jurisdictional issues concerning cloud storage of digital material.
Part I of this article considers the terminology of the search warrant and the intersection of search warrants and electronic material. The latter is described by the English Law Commission in its 2018 report as ‘the area which is in most need of reform’ 2 because the myriad different search warrant provisions 3 largely pre-date the use of electronic devices and the various methods of storing electronic data.
Part II covers filter protocols and the use of IT teams, independent lawyers, ‘taint teams’ or special masters to filter out privileged material. It considers these subjects through the lenses of approaches that have been developed by specific investigative or prosecuting authorities in England and Wales, New Zealand and the United States.
In Part III, the approaches identified in Part II are compared and contrasted to identify transferable good practice. Consideration is given to the optimal use of keyword searches, including the issue of keywords that may reveal privileged information, and the desirability of legislation regulating the choice of filtering method.
Informed by the comparative research described in the earlier sections, the article concludes with significant recommendations for key measures that are not jurisdiction specific. The measures recommended are relevant to inform the development or enhancement of legislation, regulations or codes of practice in any jurisdiction in which there is potential for legal professional privilege (or an equivalent concept) to be infringed in the context of search and seizure of digital material.
Part I
Drafting the Search Warrant
The execution of a search warrant has the potential to infringe Fourth Amendment rights in the United States, rights safeguarded under the Human Rights Act 1998 in England and under the Bill of Rights Act 1990 in New Zealand. There have been Law Commission consultations relating to search warrants, which have included consideration of electronic and privileged material in New Zealand in 2017 and in England in 2020. In the United States, the DOJ has issued guidance relating to both searches of computers and obtaining electronic evidence. Ameliorating the risks of over-seizing leads to consideration of the optimal manner in which to draft and execute search warrants. The key is to minimise the possibility that privileged material will inadvertently be disclosed whilst allowing relevant law enforcement agencies to carry out their functions effectively.
When drafting statutory powers to allow the search and seizure of information held in electronic format, the terminology used to cover all possible formats and locations of information that are likely to be searched requires some thought. Harvey notes the difficulties faced by legislators to try and encompass searching and seizing intangible data, which is not at that point in any immediately readable format, where the ‘ultimate fruit of the search will be the representation of the information in comprehensible format…’ 4 The search and seizure process itself is also impacted where digitally stored information is involved. Kerr noted that, where digital information is involved, the traditional ‘one-step search process is replaced by a two-step search process’ 5 because computer hardware will often be seized and taken off site for data on it to be copied. This copy will then be searched at a later place and time with relevant material, minus any bookmarked data, handed to the investigating team. This raises a question relating to the framing of the search warrant and whether the power to search and seize is related to the device itself (such as the hard drive, tablet and mobile phone) or the information on the relevant device. As is seen below, this conceptual challenge has been grappled with by the New Zealand Law Commission and the Law Commission in England, as part of wider consultations on search warrants, and by the DOJ in the United States. The choice of device or data may also have implications for the way in which a search takes place and how privileged material is protected.
In general terms in the three jurisdictions considered in this article, search warrants must identify the place to be searched and must identify the articles to be sought in reasonable detail. 6 Drafting a search warrant where material in electronic form is present gives rise to the choice of specifying whether the warrant gives the power to search relevant electronic devices at the specified premises/location or gives the power to search for relevant data at the specified premises/location. A search for devices involves utilizing the ‘single item theory’. 7 This approach and its ramifications have been considered by both the English and New Zealand Law Commissions. The English Law Commission identified a line of cases where the electronic device and the data stored on that device are viewed as a single item, like a book, rather than being viewed as a container of numerous separate items compared to a filing cabinet. 8 This approach allows electronic devices found at the premises specified in the warrant, such as mobile phones, computers, laptops and iPads, to be searched for and seized, even though each device is likely to contain a significant amount of irrelevant material. It also avoids having to draft a warrant which identifies which material on the device is being sought. 9 In its 2020 report, the English Law Commission identified that a warrant to search premises for electronic devices was far more popular with investigators than a warrant to search and seize data. 10
In its 2017 review of the Search and Surveillance Act 2012, the New Zealand Law Commission/Te Aka Matua o te Ture grappled with issues around the detail required in a warrant where electronic data and electronic devices are involved. 11 The Commission considered provisions in the Search and Surveillance Act 2012 (the Act) which they regarded as allowing an electronic device to be searched in the same way as a physical receptacle 12 (treating the electronic device like a filing cabinet) with no requirement for further authorisation. This would be the case where the warrant related to a place or vehicle rather than a thing and the device was present in that place or vehicle. 13 The Commission felt that this physical receptacle approach to searching devices found in a place or vehicle was no longer appropriate, primarily because electronic devices can store such significant amounts of data – far more than any physical object such as a filing cabinet (an issue that had been considered by the New Zealand Supreme Court in Dotcom v Attorney-General 14 ). The Commission described the physical receptable approach as ‘outdated’ 15 and concluded that the statutory powers in the Act 16 should not be seen as permitting an electronic device to be searched where that device is not identified in the warrant. 17 In its report, the English Law Commission suggested changes to clarify what is permitted under the relevant warrant. In relation to devices, the warrant should first specify the device to be searched for and seized and secondly should specify the information on the device that is sought. 18
In the United States, the device or data approach depends upon whether probable cause relates to a device as, for example, an instrumentality of crime, or whether probable cause relates to data on a device. 19 For probable cause to relate to the device itself, it must be more than a container – it must for example be ‘an instrumentality of the crime’. 20 If probable cause relates to information on a device that may be evidence of crime, the warrant should specify that it is data being seized and what that data is with some degree of particularity. 21 For example identifying ‘records that relate to a particular crime and to include specific categories of the types of records likely to be found’. 22 In considering the Fourth Amendment protection from unreasonable search and seizure in the context of the huge amount of data that can be stored in digital form on one device, the focus of the courts’ is ‘more on ‘how’ the agents carried out the search’ rather than what the warrant allowed agents to search. 23 Though such warrants ‘must be examined to determine if they are overly broad’. 24
It has been acknowledged that regarding electronic devices as single items does lead to the greater likelihood of privileged material being seized. 25 This likelihood needs to be considered in the context of the relevant legal framework covering the protection of privileged material which is considered in detail in Part II. 26 Clearly, in certain circumstances, seizing an entire electronic device increases the possibility of privileged material being seized.
All three jurisdictions recognise the challenges posed by large amounts of data in digital form being held on one, small, device. The ‘easiest’ approach for investigators, as evidenced in the English Law Commission report, is likely to be simply to draft the warrant in terms of searching for devices at the relevant premises rather than specifying what data is being searched for on a particular device. The English warrant application template 27 includes describing what devices are being searched for, although it doesn’t explicitly cover the information on the device that is being sought. Both the English and New Zealand Law Commissions have proposed that greater detail should be included in the warrant in terms of identifying the devices that are to be searched for at the relevant premises. Best practice would suggest that some detail regarding the information on the device that is being sought should also be included. Whilst drafting warrants to reduce the risk of unnecessary seizure of devices will reduce the risk that privileged digital material will be seized, a key issue is the adoption of processes that facilitate the sifting of privileged information from the material seized without that information being disclosed to investigators or prosecutors. These issues are considered below.
Part II
Safeguarding Legal Professional Privilege When Seizing and Sifting Digital Material
In evaluating processes for safeguarding legal professional privilege in circumstances in which electronic material to be seized may include privileged material, the obvious reference point is the rule of law rationale that underpins the existence of legal professional privilege. 28 This is reflected in the protection given to privileged material in each of the three jurisdictions considered in this article. If clients believe that a likely outcome of such processes is that privileged materials will come into the hands of those tasked with investigating or prosecuting relevant offences, the risk is that they will be less likely to confide in their legal advisers. The result is that the rule of law rationale, which regards the obtaining of legal advice as being in the public interest, will be undermined. Conversely, if such processes are transparent and their nature is such as to reassure clients that privileged material will not fall into the hands of investigators or prosecutors, clients should feel able to communicate in confidence with their legal advisers and the privilege's rationale will be upheld.
Such safeguarding processes could take a variety of forms. They could be designed to prevent initial seizure of privileged electronic material or to optimise the sifting of such material after it has been seized. They might involve different levels of human or machine-based interventions such as the deployment of non-lawyers (e.g., digital forensics specialists) using search technologies such as keyword search terms. They might involve the use of independent lawyers at a preliminary stage or at a later sifting stage following initial sifting by others and protection via deployment of ethical walls.
The Position in the English Legal System
In England and Wales, Annex A to The Attorney General's Guidelines on Disclosure 29 provides that, other than where it is seized under the seize and sift powers created by Part 2 of the Criminal Justice and Police Act 2001, digital material should not be seized if there are reasonable grounds to believe it is subject to legal privilege. 30 College of Policing guidance indicates that where the presence of privileged material on premises is anticipated, the search warrant application should make this clear and explain the safeguards that will be implemented to prevent it being seized, which may involve the instruction of independent counsel to attend in an advisory capacity. 31 Bar Council guidance recognises that a barrister may be instructed as independent counsel to attend the search either where investigators believe that there may be privileged material or because privilege has been claimed. 32
Where seizure of material which may include privileged material does take place under Part 2 of the 2001 Act (i.e., because it is not reasonably practicable to separate the privileged material from the other material on the premises 33 ), College of Policing guidance recognises that whilst it is lawful for the officers to seize the material, it should be examined and separated by independent counsel. 34 Annex A to The Attorney General's Guidelines on Disclosure provides that privileged material and material which it is suspected may be privileged (including unanticipated material discovered when reviewing material) must be isolated from other seized material 35 and members of investigative or prosecution teams should not have access to privileged material. 36 If ‘search terms or other filters’ are used to search voluminous material for privileged material, this must be ‘done by someone independent and not connected with the investigation’. 37 If material is identified as potentially privileged, there must be a review by ‘a lawyer independent of the prosecuting authority’. 38 Those who deal with privileged material must record how the privileged material was handled, who had access to it and what decisions were made in relation to it. 39
The element of Annex A that would be most likely to cause concern to clients, on the basis that it might seem to have potential to undermine the privilege's rationale, would appear to be the potential involvement of persons employed by investigative authorities, operating behind ethical walls, in the sifting process. The exemplar of the Serious Fraud Office (SFO)'s approach is instructive in this regard. In Regina (McKenzie) v Director of the Serious Fraud Office, 40 the process adopted by the SFO was to deploy its own technical staff, independent of the investigation team, using search terms provided by the owner of the devices or their legal advisers, to search seized devices for privileged material. The IT staff comprised: a team which downloaded the material onto a system called Autonomy, with the material being quarantined if thought to contain privileged content; a second team which applied search terms, with material caught by the search terms ending up in a specific folder; a third team which had expertise in the Autonomy system and ensured that the material had been uploaded successfully. The former two teams were internal to the SFO but were not part of the investigation team, the third team being external to the SFO. Whilst some members of these IT teams could access the quarantined material to see if material had been uploaded successfully, this did not involve reading the quarantined material (there was no obvious reason why they would want to read it unless they wanted to act improperly). The IT process of quarantining material was followed by review of the quarantined material by an independent lawyer, with the investigation team only being given access to material that the independent lawyer found not to be privileged. Upon the facts of McKenzie, the solicitors acting for the owner of the devices asserted that the SFO's process was unlawful and refused to provide the search terms that the second IT team needed.
The original 2011 version of Annex A (in the original para 31, a provision that was not replicated in subsequent versions of Annex A) provided that, to determine whether it is privileged, material may be examined by ‘a person independent of the investigation’, who may be employed by ‘an investigative body’ but must not be connected with the relevant investigation.
41
The Divisional Court held that even though the former para 31 had been removed from the 2013 version of Annex A, the process followed by the SFO was ‘not inconsistent’ with Annex A and in any event Annex A was not binding on the SFO.
42
More fundamentally, the Divisional Court held that: ‘a seizing authority has a duty to devise and operate a system to isolate potential LPP material from bulk material lawfully in its possession, which can reasonably be expected to ensure that such material will not be read by members of the investigative team before it has been reviewed by an independent lawyer to establish whether privilege exists. That approach to LPP material imports the necessary rigour required by the law for its protection in this context…’.
43
‘[t]here is a world of difference between determining whether something is protected by LPP, which involves close consideration of the content and context of a document or communication, and identifying a document, file or communication as potentially attracting LPP, which does not. As is plain from the description of the system for separating electronic files by the use of search terms there is in fact no need for the electronic file to be viewed at all’. 45
This former process, unlike the latter, will require consideration both of whether the prerequisites of a valid claim of privilege are satisfied 46 and, in appropriate circumstances, of whether the crime-fraud exception 47 is operational.
The court also indicated that guidance should be in place to deal with the situation in which an investigator inadvertently reads privileged material and that it should make provision for recording and reporting of the incident and for prevention of the information being used in the investigation. This might require removing the investigator who read the material from the investigation. 48 These were matters of which the SFO was aware. 49 The court held that the SFO's process for quarantining material for review by an independent lawyer was lawful. 50
For an approach such as that adopted by the SFO to align with the rule of law rationale rather than catalyse its erosion, it is necessary for clients to have confidence in the efficacy of the ethical wall put in place by the investigative authority (including steps put in place to remedy those circumstances in which privileged information erroneously falls into the hands of an active investigator). It is also necessary for clients to have confidence in the role of the independent lawyer. The best current guidance concerning the nature of this role is that which the Bar Council has for ‘Barristers instructed as “Independent Counsel” to advise upon legal professional privilege in relation to seized material’.
51
It recognises that independent counsel may be instructed to perform their functions either when the warrant is executed or at a later stage, such as when material that has been seized is to be sifted.
52
A key element of this guidance (so far as alignment with the rule of law rationale is concerned) is that independent counsel should: ‘ensure that [they] are instructed on the basis that [they] would not be permitted to divulge to the investigators or those who instruct [them] any information seen by [them] or disclosed to [them] in the course of [their] duties that is subject to the privilege of someone other than the investigating agency’.
53
The Law Commission recommended that where digital material is seized under a search warrant its treatment should be governed by statute accompanied by a Code of Practice. 60 It recommended that statute should require investigators, within a reasonable time frame after seizure, to inform persons with an interest in the material (if they requested this information) of the following matters. What was seized, what action was taken on premises regarding electronic devices and how the digital material will be examined. 61 It recommended that statute should require investigators to return devices and return or delete digital material as soon as reasonably practicable. 62 It also recommended that persons with an interest in digital material should be entitled under statute to apply to a judge to resolve disputes regarding the investigator's intended method for examining digital material, the return of electronic devices or the return or deletion of data. 63 The Law Commission did not recommend pre-search protocols, preferring the use of protocols developed post seizure in the context of the other safeguards that it recommended in its proposed statutory/code of practice regime. 64
It was recognised by the Law Commission that investigative agencies require the flexibility to adopt a process regarding the instruction of independent counsel that works for them in the context of a specific investigation (e.g., as regards whether independent counsel should be instructed to conduct an ‘on-site review’ or ‘post-search reviews’). 65 It believed, however, that guidance for such agencies to assist them in tailoring instructions to independent counsel to meet the needs of a specific investigation should be incorporated in ‘a new Code of Practice governing investigations involving electronic material’. 66 It recognised that ‘[i]n modern, document-heavy criminal investigations’ material will mostly be electronic and where hard copy material is seized the likelihood is that it will be scanned and uploaded to be reviewed electronically. 67
Concerning the overall content of the new Code of Practice, the Law Commission, whilst not intending to prescribe its overall content, recommended that the guidance it provided (which could be updated as technology and digital forensics methods changed) should include the following matters. Integrity of evidence and chain of custody, cooperation between investigators and defence teams, resolving disputes out of court, applying for warrants for digital material, the conduct of searches, the formulation of post-search protocols, examining devices and the instruction of independent counsel. 68
Unlike the position in New Zealand, the current English statutory regime does not contain any specific additional safeguards where a warrant is issued to search a lawyer's premises. The application for search warrant form, at 3(e), provides that where material (stored electronically or otherwise) may consist of or include privileged material, the applicant must specify the arrangements that will be made to sift the material. 69 Clearly, a warrant to search lawyer's premises will inherently fall within this category and the warrant will therefore have to contain very clear provisions regarding safeguarding privileged material. The Law Commission did not consider a specific regime for searches of legal premises in either its consultation or its report.
In order to identify current good practice from English Law regarding the treatment of potentially privileged digital material seized under a search warrant, it has been necessary to consider a number of sources: the Attorney-General's Guidelines, the guidance provided by the Divisional Court in McKenzie and the practices adopted by the SFO to which it refers, the College of Policing's guidance and the guidance provided by the Bar Council. The production of a Code of Practice, as recommended by the Law Commission, could bring current good practice from England and Wales and that identified from other jurisdictions together in one place, providing clear and valuable guidance for investigators and prosecutors.
The Position in New Zealand
In New Zealand, the Search and Surveillance Act 2012, ss.142–147, cover the effect of privilege on search warrants and search powers. One of the stated purposes of the Act was to modernise the law to take account of technological advances relevant to search and surveillance. 70 Legal professional privilege is recognised in s.136. In New Zealand, there is a differential regime where searches of lawyers’ premises or material held by lawyers is concerned, 71 in that a search warrant may not be executed unless the lawyer holding the privileged material (or their representative) is present. Prior to the search taking place, the lawyer must have the opportunity to claim privilege on behalf of the client. The provisions applying to searches other than of a lawyers premises/of material held by a lawyer require that a reasonable opportunity to claim privilege must be given if, having exercised a search warrant, there are ‘reasonable grounds to believe that anything discovered in the search may be the subject of a privilege …’. 72 Provisions relating to copying and securing intangible material are found at section 146 of the Act, including the process for involving the court to determine a claim of privilege. There are also requirements to detail any claims of privilege by provision of a particularised list. 73
The New Zealand Law Commission has considered whether additional special procedures ought to be in place in circumstances where computers are being searched, whether at a lawyer's premises or through remote access. For example, whether an independent expert should search any forensically copied material, perhaps under the supervision of a judge or independent lawyer. The Commission concluded that such additional measures were not necessary, because forensic copying does not involve any disclosure of the copied material and because the measures proposed to preserve privilege (as outlined above) will be equally efficacious during a search of a forensic copy. 74 The latter would require that the lawyer is given the opportunity to assert privilege prior to the search taking place, although at this point the material may no longer be at the lawyer's premises. If the warrant relates to material held by a non-lawyer, it can be executed without that person being given an opportunity to claim privilege and in fact without that person being present. For a forensic copy of intangible material, this therefore means that the copy can be searched without there being any opportunity to claim privilege, until such time as the person conducting the search has reasonable grounds to believe that material discovered might be subject to privilege.
The difficulties presented by computer searches conducted by an enforcement agency, where physical separation of intangible privileged material risks the integrity of the data, was considered by the Law Commission. 75 The Commission did not favour the use of an independent computer expert. The expertise of enforcement agencies in computer searches, the risk to the integrity of an investigation of clandestine searches of privileged material and the benefits of the search being conducted by a person with considerable knowledge of the investigation were all considered to mitigate against recommending the use of independent experts. 76 Also considered but discarded was the requirement for an initial inspection of a lawyer's computer (preliminary screening), simply to identify relevant evidential material without further inspection and then determine any claims for privilege over only this material, as a means of reducing the scope of the search and the amount of material accessed. Again, the opportunity to assert privilege prior to any search taking place was deemed to be a better approach than this, although there was a suggestion that computer searches could take place in stages, with the lawyer having the opportunity to claim privilege at each stage. 77
In its subsequent issues paper in 2016, Review of the Search and Surveillance Act 2012, the Commission identified the use of digital forensic units within enforcement agencies. These units are used where searches deal with large amounts of digital material, to capture and search this material, sending relevant material to investigators and ensuring that privileged material is not sent to investigators. 78 It noted in relation to the requirement on the person claiming privilege to provide a particularised list that in the context of digital material the practice was often, via discussion with the owner of the material, to produce a customised plan to identify and isolate privileged material. 79 This could include provision of search terms by the owner of the material, a lawyer for one of the parties identifying the materials of use to an independent lawyer or appointing an independent person to identify the material. 80
The exemplar of procedures adopted by the Inland Revenue Department is valuable and the Commission paid particular attention to them. 81 The Inland Revenue Department has a Digital Forensics Unit (DFU) comprising computer forensics staff that operates independently from the investigators. 82 Normally the DFU will access, search and copy digital material but where the DFU cannot attend on site either external staff will be contracted in or investigators will remove the devices (maintaining a clear custody chain) for delivery to the DFU. 83 The DFU does not release electronically stored documents to the investigators until the owner of the material has had an opportunity to seek legal advice and make particularised claims, the owner of the material being entitled to provide the DFU with keywords for use in identifying privileged documents. 84 If documents are identified via this key word search process the owner of the material is then given an opportunity to claim privilege in them, 85 which in the case of a dispute between the person claiming privilege and the Commissioner may ultimately end up before a district judge. 86 The Commission recognised that DFU's keep a record of the search procedure followed during a search, but did not know whether enforcement agencies lacking a DFU followed the same practice. It suggested that this practice could be required by statute. 87 It made other suggestions as follows. Statute could require the ‘issuing officer’ 88 to consider whether conditions should be imposed on a search warrant for the purposes of restricting how much irrelevant material is viewed and ‘preventing investigators from seeing privileged material’. 89 Conditions regarding the treatment of digital material could impose processes similar to the procedure followed by the Inland Revenue Department. 90 And legislation could provide that persons searching digital material should ‘take all reasonable steps to avoid seeing’ privileged material 91 (though it believed that the DFU process already achieved this goal). 92
Good practice on the part of the Inland Revenue Department's DFU as identified by the New Zealand Law Commission seems largely to align with that adopted by the SFO in England and Wales in utilising a technical team independent of the investigatory team to access, search, copy and sift digital material, including cooperation with the privilege holders in devising keyword searches. The main difference appears to be the use of an independent lawyer to review the quarantined material in England and Wales, with a judge being relied on to resolve privilege disputes in New Zealand.
The Position in the United States
The issue of how to protect legal professional privilege when material is seized under a search warrant is one which has produced much higher volumes of academic debate and judicial comment in the United States than has been the case in England and Wales or New Zealand. The American courts are divided as to how to protect privileged information seized during the execution of a search warrant. In the past most courts routinely approved the use of taint teams (also known as filter teams or privilege teams), comprised of prosecutors and agents not involved in the pending case, to review the seized materials and identify unprivileged materials to be turned over to the prosecutors in the case. They are often used when large quantities of electronically stored documents are seized. 93 Apart from conducting privilege reviews, filter teams can also be used to conduct searches on premises where there may be privileged material (e.g., where lawyers are located), interview witness (such as lawyers) who may reveal privileged information, examine covert recordings that may reveal privileged information, liaise with the defense regarding the treatment of potentially privileged material and litigate issues of privilege. 94 Best practice as regards the use of filter team personnel at the search stage is to assign a filter agent, with a filter attorney also available to be contacted, in the context of a search where potentially privileged material is reasonably likely to be present. 95
The use of a privilege team is one of three options identified in the Justice Manual where the premises of an attorney who is a subject of an investigation are searched.
96
Today many courts are suspicious of taint teams. These courts believe that there is a risk to the privilege – trusting a taint team comprised of prosecutors is like allowing the fox to watch the ‘henhouse’.
97
That concern certainly has merit; psychological research has underscored the effect that even subconscious bias can have on decision-making in the criminal justice system.
98
Even if at a conscious level a prosecutor or a taint team member strikes mightily to make an objective, impartial, decision, at a subconscious level he or she may be tempted to find that a particularly damning letter or email is unprivileged. Other dangers include ‘vague procedures’ that allow documents to ‘fall through the cracks’ and ‘intentional misconduct’.
99
Fundamentally, ‘the taint team and prosecution team are in the same organization, report to the same leaders, and often share the same goals’.
100
Department of Justice guidelines provide that: [w]hen agents seize a computer that contains legally privileged files, a trustworthy third party must examine the computer to determine which files contain privileged material. After reviewing the files, the third party will offer those files that are not privileged to the prosecution team. Preferred practices for determining who will comb through the files vary widely among different courts.
101
A special master has been described as ‘a disinterested third party [who] could be anyone from a judicial officer, an attorney, a law professor, to a former attorney for the DOJ’ and other examples have included ‘a former judge’ and ‘a United States Magistrate Judge’, both appointed as special masters. 109 Frohock suggested that ‘When available, a judge or Special Master provides the advantage, at least on optics, of a knowledgeable and neutral eye, bringing independence from outside the executive branch’. 110
As an exemplar of the approach to filtering privileged material in the United States, the DOJ's Criminal Division now has a Special Matters Unit (SMU) which has developed procedures to allow operation across Federal districts. Before considering the processes followed by the SMU, it is first important to consider recent judicial restrictions on the use and operation of taint teams as the SMU's processes are intended to take account of such case law. In particular, two recent federal decisions – one from Court of Appeals for the Fourth Circuit and another from the Eleventh Circuit – have called past practice concerning the use of taint teams into question and recognised that when they are used ‘exacting requirements’ 111 need to be satisfied. The first, but not the second, concerned a search of lawyers’ premises.
The Fourth Circuit opinion, In re: Search Warrant Issued June 13, 2019, 112 provides a road map for lawyers who are faced with the use of a filter team. The case concerned an allegation that Lawyer A (a partner in a twenty-partner law firm) had assisted drug dealers with criminal activities such that the crime-fraud exception was applicable. An Internal Revenue Service (IRS) agent obtained a warrant to search the relevant law firm's offices. The warrant was obtained at an ex parte hearing in which the defense could not participate. The warrant authorised a filter team, which was not involved with the investigation, and included lawyers from another U.S. Attorney's office, a legal assistant and a paralegal who also worked there, agents of the IRS and of the Drug Enforcement Administration, and forensic examiners.
The process, proposed by the government in a Filter Protocol attached to the warrant application and approved by the magistrate judge, allowed the filter team to determine, in the first instance, which materials were potentially privileged and which were not. It permitted the team to turn over to the investigative team those documents which it identified as non-privileged without any court determination. If the filter team identified documents as privileged and ‘responsive to the search warrant’ it was required to categorise them as: incapable of redaction; capable of redaction; or potentially privileged (e.g., where the crime-fraud exception might be applicable). In relation to the latter two categories, if agreement could not be reached with Lawyer A's counsel, the matter would be determined by the court. If the client waived privilege, the documents could be given to the prosecution team without additional review.
In executing the search warrant, the agents seized emails to Client A, and about 50,000 other emails. Many of these related to other clients and attorneys. The law firm applied for a preliminary injunction, which was granted on appeal. In the course of explaining its decision, the court identified a number of errors that the magistrate judge had made when approving the filter team and the Filter Protocol. The authorisation of the filter team and Filter Protocol by the magistrate judge had delegated a judicial function to the executive. This was compounded both by the fact that non-lawyers such as agents and paralegals were authorised to categorise documents as non-privileged and the risk that filter teams might make errors of judgement both in privilege determinations and in the transmission of privileged material.
The authorisation of the filter team via an ex parte hearing prior to the warrant being executed meant that the magistrate judge had not been fully informed when he made his decision. For example, the timing issue meant that the judge could not have known that only a tiny percentage of the emails seized was from or to client A (with many concerning other clients). And had the hearing been adversarial, the magistrate judge would have known more about the law firm, the firm's clients and the seized materials.
In contrast to the flawed procedures employed by the magistrate judge, the court endorsed the ‘sensible’ 113 procedures used in another case, 114 when agents searched the law office of Michael Cohen. In Cohen, adversarial proceedings re the proposed filter team had taken place four days after the materials were seized and prior to their review by the filter team. Rather than authorising the filter team the court appointed a special master. The procedures followed in Cohen had the following consequences. 115 No seized items were reviewed by Executive Branch personnel prior to a privilege review being conducted. The court appointed a special master, a retired United States District Judge, to perform the privilege review. Relevant clients were allowed to intervene and participate in the privilege review process. The seized items were provided to the lawyer and relevant clients, who were given an opportunity to mark items as potentially privileged and/or ‘highly personal’. Items not marked as potentially privileged by the lawyer or relevant client(s) were provided to the Executive Branch. Items marked as potentially privileged or otherwise confidential were reviewed by the special master. The special master regularly conferred with the lawyer and relevant client(s) about the proper designation for items as to which she had questions or required further argument. The special master provided written reports and recommendations to the court on a rolling basis regarding the special master's proposed privilege determinations. The lawyer and relevant client(s) had an opportunity to object in writing to the special master's reports and recommendations. And the court conducted an independent review of the reports and recommendations and the documents in question before issuing any rulings on the privilege determinations. 116
The Fourth Circuit also identified a failure on the part of the magistrate judge and of the district court to ‘explicitly weigh those foundational principles that protect attorney-client relationships’, 117 with the result that ‘the Filter Protocol authorized government agents and prosecutors to rummage through Lawyer A's email files’, many of which ‘concerned other clients and other matters’. 118 The court did not totally preclude the use of filter teams. Rather, it issued a more limited ruling that under the circumstances the filter team and protocol were improper. 119
In contrast to a law office search, the Eleventh Circuit Decision in In re Sealed Search Warrant and Application for a Warrant by Telephone or Other Reliable Electronic Means 120 involved the search of a suite of family businesses. The items seized included the documents of in house counsel. The search warrant authorised, inter alia, the seizure of emails and imaging or copying of computers or other storage media. The Filter Protocol required that if communications with an attorney were identified by the Government, the investigative team would cease its review until a filter team (of government attorneys and agents who were not involved in the investigation) had segregated communications to which the attorney-client privilege might attach. The filter team would pass on non-privileged communications to the investigative team but would not inform it of the content of communications with an attorney. If it decided that communications with an attorney were not privileged, it could only pass them on if it obtained a court order.
After the search, the owners, managers and controllers (collectively, the ‘Intervenors’) sought injunctive relief to prevent the filter team from reviewing communications that might be privileged unless the intervenors agreed, or the court reviewed them for privilege and ordered their disclosure. In the district court, the magistrate judge imposed what is described throughout the Eleventh Circuit's decision as a ‘Modified Filter-Team Protocol’. Under this Modified Protocol, the initial privilege review was to be carried out by the Intervenors, who would identify communications in which privilege was asserted in a ‘privilege log’ and release to the investigative team items which were not privileged. The filter team could not contain staff from the branch office that was conducting the investigation, it could not share a ‘first level supervisor’ with any investigation or prosecution team members and filter team supervisors were to be ‘walled off’ from the investigation. The filter team could review items in the privilege log and challenge the Intervenors’ assertions of privilege. The filter team and Intervenors would liaise to try to resolve such challenges. In the absence of resolution, a dispute would be resolved by a ruling from the court or a special master. The filter team would release communications to the investigative team if the parties agreed or privilege was overruled by the court.
On appeal, the Eleventh Circuit, in upholding the Modified Protocol, recognised that, under it, the Intervenors were entitled to carry out the initial privilege review and that the filter team could not inadvertently release privileged communications because either the Intervenors permission or that of the court was required before allegedly privileged documents could be released. It gave three reasons for approving the Modified Protocol. First, some other circuits had approved the use of filter teams. Secondly, no authority had been identified for the Intervenors’ proposition that review by government agents of allegedly privileged communications should never take place until a court had determined whether they were privileged. Thirdly, the Modified Protocol did not exhibit any of the defects that had resulted in specific protocols being disapproved by other courts.
In deciding that the Modified Protocol ‘compl[ies] with even the most exacting requirements other courts that have considered such protocols have deemed appropriate’, the court distinguished the position in In re: Search Warrant Issued June 13, 2019 121 in which, as was seen above, the Fourth Circuit had not approved a Filter Protocol. In the case before the Eleventh Circuit, an adversarial hearing had taken place before the Modified Protocol was implemented. There was no suggestion that the bulk of the privileged communications seized was irrelevant to the client under investigation. And, rather than delegating judicial functions to the Executive, initial privilege review was conducted by the Intervenors with allegedly privileged communications only released to the investigative team if the Intervenors agreed or the court so ruled.
In the United States, ‘[t]here is no uniform, nationwide standard for conducting filter reviews, and there is no one size-fits-all protocol that a filter team should follow’. 122 The DOJ's Criminal Division now has its SMU, ‘an independent, specialized, in-house filter team’, established in 2020. 123 Taking into account recent case law concerning the use of taint teams, it has developed procedures which are intended to be applicable across multiple Federal districts. 124 In relation to digital evidence which may include privileged material, its ‘“Attachment B” template’, which is included in search warrant applications, articulates the following process. 125 The SMU ‘will conduct an electronic review of the data’ using ‘privilege search terms’, compiled ‘in consultation with the search team’, which ‘include specific names and generic words intended to identify potentially privileged information’. 126 If this process identifies material as ‘not potentially privileged’ it may be ‘released to the search team’ whereas where it identifies material as ‘potentially privileged’ it is segregated for review by an SMU lawyer who determines whether or not it is ‘potentially privileged’. 127 If the lawyer determines that it is not it may be released whereas if it is the lawyer has three options. First, make an ex parte application to the court for a determination. 128 Secondly, defer making such an application and ‘segregate the documents in a manner that makes them inaccessible to the search team’. 129 Thirdly, disclose them to the party entitled to claim privilege, requesting a privilege log if privilege is asserted and apply to the court for a ruling if an agreement is not forthcoming. 130 In the course of the filter process, the SMU does engage with privilege holders ‘where practicable’ and in ‘many cases’ the protocols are agreed by the privilege holder or approved by the court. 131 This process was designed to deal with criticisms made by the courts in cases such as In re Search Warrant. 132
The SMU has been criticised on the basis that it is integral to the investigative and prosecuting authority and works with the search team. 133 Frohock argues that whether or not the Unit ‘makes sense’ depends on whether prosecutions are adversarial or quasi-adversarial. 134 It makes sense to Frohock (and proceedings are better categorised as quasi-adversarial) if prosecutors, seeking ‘justice rather than convictions’ are ‘cast as ministers of justice’. 135 If they are not so cast (and proceedings are better categorised as adversarial) then, Frohock asserts, ‘every in-house privilege strategy rings hollow’. 136
Even if prosecutors do strive to act as ministers of justice, this does not remove the possibilities, identified above, of subconscious bias in decision making or of procedural defects that result in unintended disclosure. More fundamentally, it does not remove the likelihood that privilege holders will assume that prosecutors will not act as ministers of justice, thus undermining the privilege's rationale if it results in reluctance to take legal advice. The practices adopted by the SMU and the approach of the Courts in the two US cases considered above are compared in Part 3 to those that were identified above from England and Wales and New Zealand.
Part 3 Identifying Transferable Good Practice
The Processes Adopted by the SMU, the SFO and The Inland Revenue Department (and its DFU) Compared
There are significant differences between the process delineated by the SMU's template and those adopted in England and Wales by the SFO, as approved by the Divisional Court in McKenzie and by the Inland Revenue Department's DFU in New Zealand. Under the SFO's process as articulated in McKenzie, the search terms were provided by the privilege holders, the IT process via which the search terms were applied to the electronic material was conducted solely by IT staff (who did not read the material) and quarantined material was reviewed by an independent lawyer, not employed by the SFO. Similarly, under the Inland Revenue Department's processes in New Zealand, the DFU receives key words from the privilege holders who are entitled to claim privilege in any material identified by this process. In contrast, the SMU team contains agents and lawyers who, unlike the IT experts deployed in the other two jurisdictions, are employees of the investigative/prosecution authority. Whilst not forming part of the investigative team, they are likely to understand the privileged material and its significance to the investigation. This would seem to suggest that when compared to the approach adopted by the SMU, both the SFO's process and that adopted by the Inland Revenue Department might reduce significantly the likelihood that investigators or prosecutors may, by accident or design, be able to access privileged material and/or that decisions concerning the privilege status of quarantined documents may be contaminated by subconscious prosecutorial bias.
It is informative to review these respective processes through the lenses provided by the decisions In re: Search Warrant Issued June 13, 2019 and In re Sealed Search Warrant and Application for a Warrant by Telephone or Other Reliable Electronic Means. Provision of search terms by the privilege holder and review of the content of documents by an independent counsel under the SFO's process would appear to come closer to an approach that the Fourth and Eleventh Circuits might have preferred (in involving the privilege holders in the process) than the SMU's process under which search terms are the product of consultation between the SMU and the search team (even though the filter team will engage with the privilege holder ‘where practicable’), and the documents are reviewed by a filter team lawyer. Independent counsel is more akin to a special master than is a filter team lawyer, even though independent counsel is instructed by the SFO, not by the court. An issue with the SFO's process that might have troubled the Fourth and Eleventh Circuits is whether independent counsel would have been permitted to communicate with the privilege holder (whether this is permissible is something which under the Bar Council's Guidance the instructions given to the independent counsel should make clear). This could also be a problem with the SMU's process. When review is by one of its lawyers and the lawyer determines that the material is not potentially privileged, this review may take place in circumstances where it is not ‘practicable’ to engage with the privilege holder before disclosing to the investigators. 137 The SMU has recognised, however, that ‘filter teams [should] remain cognizant of the state of the law and proactively model their processes to reflect court guidance and anticipate future challenges’. 138 The Inland Revenue Department's process does not involve an independent counsel, but a claim of privilege may ultimately end up before a district judge.
The Use of Keyword Searches and the Potential for Legislation Regulating the Choice of Filtering Method
In terms of identifying privileged material and keeping it from investigators and prosecutors what is the ideal combination of technology and humans? Adams suggested that ‘search and privilege doctrine have evolved much more slowly than technology’ 139 but technology also has its limitations. Whilst the authorities considered in this article suggest that keyword searching is the norm where electronic communications which may contain privileged content have been seized under a search warrant, in future it may be that the use of other filtering methods, such as predictive modelling, 140 should also be considered. From the context of disclosure/discovery in civil proceedings, it seems that keyword searching can perform better than predictive modelling but that the opposite can also be true and that using predictive modelling to compliment keyword searching increases the level of precision. 141 The precision of document review is significantly increased if predictive modelling is supplemented by subsequent manual review. 142 Too extreme a level of reduction of human input, with an overreliance on technology, increases the risk that privileged material will not be identified and will slip through the net, because manual review will identify some documents that predictive modelling will miss. 143
As regards keyword searches, selection of keywords can be crucial. If the keywords selected are too broad this may result in increased expense if lawyers are required to review an unnecessarily high volume of documents, whereas if they are too narrow this may result in privileged documents not being identified 144 . For example, in a recent decision of the High Court of New Zealand, 145 the applicant, whose devices and electronic documents had been seized by the New Zealand SFO under a search warrant, provided 67 key words which included a number of generic words, such as ‘affidavit, defense, defence, legal brief, law, lawyer, trustee, beneficiary, barrister, and legal’. 146 A search using these keywords identified 54,000 documents with an additional 500,000 documents attached to them or referenced in them. Reducing this to 24 key words in the form of the full names of lawyers and law firms and domain names (which the SFO indicated is what it would normally receive from a party asserting privilege) reduced this to 4210 documents.
The practices adopted by the SMU, the SFO and the Inland Revenue Department's DFU concerning the use of keyword searches and the selection of keywords are not imposed by legislation. Rather, as has been seen above, it is developing and enhancing practice on the part of investigative bodies, at times driven or enhanced by judicial decisions endorsing or criticising such developing practice, that has been the catalyst for change. In a recent decision of the European Court of Human Rights, Särgava v. Estonia, 147 the court found that the seizure and examination of a mobile phone and a laptop had given rise to a violation of Article 8 of the European Convention on Human Rights (the right for respect to private and family life, home and correspondence). The observations made by the European Court of Human Rights 148 included that whilst the investigative authority had provided an assurance after the devices were seized that a keyword search would be used, domestic legislation did not impose an obligation to carry out a keyword search. This resulted in a keyword search not being mentioned in the search warrant application and the warrants themselves not mentioning any such obligation. In consequence, the decision whether to carry out a keyword search or use other filtering methods and the choice of keywords was entirely for the investigatory authority and the law did not make clear how disputes about which keywords or which filtering methods to use were to be determined.
Reference to Särgava suggests that statutory provisions that authorise and regulate the issuance of search warrants might be enhanced by the inclusion of specific provisions concerning how the method or methods for filtering electronic evidence is or are to be determined. And how disputes concerning the method or methods to be used and the implementation of such methods (e.g., the selection of keywords) are to be resolved. Given the speed of technological change, however, the inclusion of statutory obligations to utilise a particular filtering method, would be less desirable. It could restrict the implementation of developments, such as blending keyword searches with predictive modelling and manual review in appropriate cases; developments which might both increase the efficiency of the filtering process and reduce the overall cost of the process. Responsiveness to technological change is one reason why the Law Commission recommended a combination of statutory safeguards and a related Code of Practice, with the latter being easier to modify in the context of ongoing and rapid technological development. 149 It recognised that sifting digital material has cost implications and can require the expenditure of ‘large volumes of resources’. 150
A common theme of the approaches adopted by the SMU, the SFO and the Inland Revenue Department's DFU is the use of personnel who are independent of the investigation itself but integral to the investigative or prosecuting authority. A recent Canadian decision (of the Supreme Court of British Columbia) preferred the use of an independent court appointed computer forensics technician to the use of a forensics team integral to but operationally independent of the investigative authority.
151
One of the risks that the privilege holder had asserted was that ‘keywords themselves can reveal solicitor-client privileged materials’.
152
The problem that the court identified was that; ‘…using specific keywords likely results in a more thorough isolation process, but may result in divulging privileged information by mistake. On the other hand, using general keywords reduces the likelihood of keywords containing privileged information but may result in incomplete review of files, an under-inclusive isolation effort, and subsequent inadvertent disclosure of solicitor-client privileged materials’.
153
It seems unlikely that the courts in England and Wales, New Zealand or the United States could be persuaded to mandate the adoption of a similar approach as a general rule (not least due to concerns regarding cost) or (and equally unlikely) that such a rule would be imposed by statute. It is clear, however, that there are circumstances in which courts in the United States will appoint a special master to replace a filter team 155 and it may be that in appropriate circumstances and on the facts of a specific case a court in England and Wales, New Zealand or the United States might prefer an independent forensics technician to one integral to an investigative team or prosecuting authority. This might be the case if, for example, discussions/information to identify the most effective keywords (maximising the likelihood of capturing privileged documents most accurately and, potentially, most cost-efficiently) could reveal privileged information.
Conclusion
In this article, the strengths and weaknesses of methods of protecting legal professional privilege in the context of seizure of digital material under search warrants in three jurisdictions have been compared, with particular reference to the practices adopted by specific authorities in each jurisdiction as exemplars. Informed by this comparative analysis, we now conclude with recommendations. We suggest that to achieve transparency and consistency and maximise the likelihood that the rights of privilege holders to effectively assert privilege will be observed, relevant processes and safeguards should be enshrined in statute with associated regulations, codes of practice or guidance to allow flexible response to technological change. Exactly what form such legislation, regulations, codes of practice or guidance should take may vary between jurisdictions, but they must be realistic and proportionate in terms of resultant workload/costs implications. Our recommendations are not intended to be jurisdiction specific but identify key elements of relevance to any jurisdiction where legal professional privilege or an equivalent concept exists and may be infringed by the exercise of search and seizure powers.
Legislation, regulations or codes of practice designed to preserve legal professional privilege in the context of seizure of digital material should address the following issues. Where a search warrant includes electronic devices, these should be described on the warrant and there should be sufficient clarity in the application about what information on the device is being sought to enable the authorising judge or equivalent to determine the likelihood of over seizing. The application should inform the judge or equivalent if there is a likelihood that digital material seized may include privileged communications. If there is such a likelihood, the application should articulate the process that will be adopted to ensure that, when seized, privileged material is filtered from other material and does not fall into the hands of investigators or prosecutors.
Where there is a likelihood that digital material will include privileged material (e.g., a law firm or in house legal department), a lawyer independent of the investigative and prosecuting authorities should be involved in the search and seizure process. The use of a lawyer employed by an investigating or prosecuting authority, even if not assigned to the case and belonging to a different unit or command, gives rise to at least the perception of bias and the possibility of subconscious bias should be avoided. The devices or material should be seized by a digital forensics team operating independently of the investigating or prosecuting teams. Digital forensics teams must be independent of the investigating or prosecuting team but can be integral to the investigating or prosecuting authority provided that the process they conduct does not involve them in reading any of the seized material. If potentially privileged material is discovered in circumstances in which it was not expected, the sifting of the material should not be undertaken by investigators or prosecutors but by a combination of an IT team operating independently of the investigating or prosecuting team and an independent lawyer, special master or judge. The IT team using technology such as keyword searches to sift provisionally privileged material must be independent of the investigating or prosecuting team but can be integral to the investigating or prosecuting authority provided that the process they conduct does not involve them in reading any of the seized material. It is crucial that the privilege holder and their legal advisers have the opportunity of claiming privilege from as early a stage of the search and seizure process as possible. This should include opportunities to communicate with an independent lawyer, where one is deployed. It should also include opportunities to input their views in the context of the sifting process, for example, when keyword search terms are being formulated. It is also crucial that any process designed to reduce the risk that privileged material will fall into the wrong hands is realistic in terms of the workload/costs implications that it imposes. This makes the selection of appropriate keyword search terms particularly important. It also means that legislation, codes of practice and guidelines should not hamper the introduction of new technologies that make the filter process faster and cheaper, provided that they do not increase, or indeed reduce, the risk that privileged digital material seized under a warrant will fall into the wrong hands.
Legislation, regulations or codes of practice should make provision for expedited judicial resolution of privilege disputes. They should also make clear that where the filter protections fail such that privileged material falls into the hands of an investigator or prosecutor, the relevant personnel are precluded from any further involvement in the investigation or prosecution.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.