Abstract – In recent years, Europe has witnessed several incidents involving high-ranking political officeholders using text or instant messages for official business and being unable to retrieve them at a later stage. Prompted by these events, the State Archives of Belgium set out to map the challenges of, and potential strategies for, safeguarding this new type of archive. This paper outlines the result of this endeavor. After a brief introduction and summary on the emergence of texting and instant messaging as pivotal communication tools, we first assess their suitability for policy-making dialogue. Then follows an exploration of the preservation challenges the medium presents, both in a general context and specifically within Belgium, and of some strategies that can be employed to mitigate these challenges. In the conclusion, we summarise and put forward additional steps that archives can take to ensure the preservation of text messages by political officeholders, and briefly contemplate on the disclosure of these messages to the public in the short and long run.

Keywords – text messages, instant messages, mobile data, political archives, government archives.

This paper was submitted for the iPRES2024 conference on March 17, 2024 and reviewed by Maureen Kenga, Bonface Odhiambo and 2 anonymous reviewers. The paper was accepted with reviewer suggestions on May 6, 2024 by co-chairs Heather Moulaison-Sandy (University of Missouri), Jean-Yves Le Meur (CERN) and Julie M. Birkholz (Ghent University & KBR) on behalf of the iPRES2024 Program Committee.

Introduction

In recent years, the use of text messages and instant messaging by individuals in policy-making positions has led to a number of notable incidents in Europe, mainly due to the inability or unwillingness to archive these communications for public scrutiny. One prominent example from early 2022 involved European Commission President Ursula von der Leyen, who was criticised for failing to disclose messages she had exchanged with Pfizer CEO Albert Bourla about the purchase of a significant quantity of COVID-19 vaccines. Later that year, the prime ministers of the Netherlands and Belgium became embroiled in similar controversies, albeit each with their own angle: while Dutch PM Mark Rutte admitted to routinely deleting text messages from his mobile phone to free up storage space, his Belgian counterpart, Alexander De Croo, was accused of withholding messages that allegedly showed his cabinet had approved a budget with errors, for which the responsible minister had previously been dismissed.

These occurrences, along with the media attention they garnered, prompted the State Archives of Belgium to undertake an exploration of the legal, political, and technical aspects of archiving such digital communications, an area it still was largely unfamiliar with. The goal was twofold: first, to understand the technical constraints and legal framework within which such information could be archived, and second, to identify steps it can take and recommendations it can make to political officeholders to ensure that this vital aspect of the nation's collective memory is preserved, along with its ability of holding elected officials—and, secondarily, the government in general—accountable in the short and long term.

Indeed, the information exchanged through mobile phones has become an integral part of the policy-making process, rendering its preservation crucial for sustaining democracy and combating mis- and disinformation. As Canadian information commissioner Suzanne Legault wrote in a special report on instant messaging to the Canadian government from 2013: “Information is the lifeblood of democracy. Loss of information or an infringement on requesters' right of access should not be quietly accepted simply because the impact of new technologies has not been properly considered.” [1]

‘Texting’ Versus ‘Instant Messaging’

In the category of mobile messaging, we can distinguish two types of messages that are widely used in political circles.1

The oldest one is SMS: short messages—140 bytes—sent via the message center of a telecom provider and containing only text—hence their nickname ‘text messages’ or ‘texts’. The first message sent via the ‘Short Message Service’ protocol was a Christmas greeting from a young British programmer to a colleague in December 1992. After that, the use of SMS grew in ten years to 250 billion texts annually worldwide.2 This success was not matched by MMS, or ‘Multimedia Messaging Service’, the protocol that emerged in 2002 and allowed the transmission of other media such as sound recordings, photos, and videos.3

This is partly due to the rise of 'instant messages' or ‘IM,’ messages sent over the internet, using an application that is not tied to one phone type or brand. The first online messaging service, BlackBerry Messenger, was launched in 2005, followed by WhatsApp—today’s most popular messaging service, which will thus be the primary focus of this paper—and many others in the years thereafter.4 Instant messaging services share many characteristics with other types of online communication services, such as social media platforms, meeting applications with chat functions such as Microsoft Teams, and internet forums. However, most IM services, like offline texting and unlike other types of online services, are mostly linked to a telephone number, entrenching their mobile nature.

Texting and Messaging In The Political Arena

Before we talk about archiving texts and instant messages, it seems appropriate to ask the question whether the use of these communications channels poses a problem, particularly from a security perspective.

Use Of Smartphones By Political Officeholders

Every so often, concerns arise regarding the security of the devices of our highest-ranking politicians and officials. Common triggers for this include the surfacing of sensitive domestic information in the hands of foreign powers or hacker collectives, or the discovery of suspicious software on the phones of politicians recently abroad. In February 2010, the Belgian State Security cautioned ministers and top officials against transmitting sensitive information via their BlackBerrys, citing a substantial risk of interception due to vulnerabilities acknowledged by the manufacturer. In 2015, the government procured a hundred secure BlackBerrys for government members and their close coworkers to further mitigate the risk.

Despite of this effort, investigations by Le Soir journalist Martine Dubuisson revealed in 2023 that most Belgian ministers, including the Prime Minister, again or still utilize a single, often private smartphone for all their communications (“les ministers pour la plupart, dont le Premier, n’ont, nous confirme-t-on à plusieurs sources, qu’un seul GSM, pas un smartphone professionel en plus d’un téléphone privé”). [2] However, this practice may soon change, as a new generation of secured smartphones, developed by the specialized Belgian Secure Communications (BSC) entity was announched in March 2024 and is set to be rolled out to all ministers, top officials, and diplomats.5

How Secure Is Texting And Messaging?

The next question that arises is the extent to which we can entrust politically and diplomatically sensitive information to telecommunications and instant messaging services, particularly when their infrastructure is located abroad or even if the companies behind them are based in other countries. For instance, the termination of the Privacy Shield agreement between the EU and the United States in 20206 has left European countries in uncertainty about the legal protection of their data stored in the US. While companies like WhatsApp also have commercial reasons to prioritize user privacy, and seem to be doing so quite successfully—a leaked FBI training document shows that even American security services can't obtain WhatsApp message contents, unless backed up in iCloud [3]—this does not provide sufficient certainty for the target group we are discussing.

When it comes to choosing between texting and using instant messaging services, neither option is unequivocally more advantageous than the other in terms of minimizing the risk of information leakage to foreign governments. The risk of transmitting messages through privately owned servers (abroad) must be balanced against the benefits of end-to-end encryption offered by most instant messaging services and not by traditional text message applications.7 Consequently, depending on the choice of phone and applications, the content of a text message might be more susceptible to interception by malicious actors than the content of a WhatsApp message or the like.

We do not have insight into the strategies adopted by Belgian ministers in making this choice. However, the aforementioned study by Dubuisson also testifies to the absence of a national-level code of good practices regarding the use of electronic messages in Belgium. Instead, reliance is placed on common sense to avoid discussing sensitive matters via both texts and WhatsApp. [2]

Archiving Texts And Instant Messages

As text messages and later instant messages made their way into all domains of life, a need for tools and policies for archiving them was soon felt, most of all in the business sphere.8 An article titled “Like E-Mail, Only Faster: Don't Forget to Address Instant Messaging in Electronic Communications Policies”, was published by corporate lawyer Jason Krause in 2005. In this prescient paper, Krause warned that “If a company allows instant messaging, which deletes messages as soon as a computer is turned off, it may need to use business-class instant-messaging software that can archive the messages. Otherwise, a company has no record if something said in an IM winds up in court.” [4] In this light, it should come as no surprise that most archiving tools for texts and instant messages are primarily aimed at the private sector. More on those later.

The public sector has appeared slower in thinking about instant message archiving.9 Still, many countries, among which Belgium, have no central guidelines on the matter. (Some of the ones that do, will be mentioned further down.) This may be attributed to the fact that archival institutions were simultaneously confronted with other challenges perceived as more pressing, such as email archiving. While emails clearly, from the beginning replaced traditional paper correspondence, text messages and instant messaging mostly came in the place of phone calls and face-to-face conversations: forms of communication that were not typically captured in the past, making their archiving initially appear more of a ‘nice-to-have’ than a ‘must-have’. However, instant messaging has now become such a significant part of communication that it has indeed become an essential link in the decision-making chain, as exemplified by the cases mentioned in the introduction.

In 2009, American archivist Michelle Caswell noted that “despite the growing importance of cell-phone-generated materials, few archivists have addressed the problems of appraising, providing access to, and preserving these materials”. [5] Her observation encompassed all forms of information created by and stored on mobile phones. At that time, Caswell noted that only two authors had written on the subject, both in response to the September 11, 2001 attacks.10 The video and audio recordings captured by witnesses on their mobile phones, along with the deeply personal text and voicemail messages sent by victims during the tragic events, prompted many archivists to contemplate, for the first time, the potential social and historical significance of information derived from mobile phones (and digital archives in general). The September 11 Digital Archive probably represented the earliest collection of such materials.

Unlike the thriving research on email, web, and social media archiving, archiving instant messaging remains a relatively marginal topic in preservation studies.11 By way of indication: in contrast to the first three issues, the Digital Preservation Coalition hasn’t yet published a Technology Watch report on archiving texts and instant messages. It has published a couple of blogs on the subject and consistently included non-standardized public records such as WhatsApp messages in its annual Bit List (Global List of Endangered Digital Species) since 2019, classifying them consistently among the most endangered archive types. [6]

Moreover, political officeholders represent a distinct group of archival producers, presenting unique challenges in the preservation of their texts and instant messages. To our knowledge, no comprehensive overview of strategies for effectively archiving such material, covering both its protection, capture, selection and long-term storage, exists. In the remainder of this paper, we aim to address this gap by providing an overview—without delving too much into technical details—of potential approaches and evaluating their suitability for the target group. For this, we draw on existing literature and experiences of other archive institutions, small-scale experiments with our own messages and a study of the automated solutions on the market.12

Preventing messages from being lost

The challenge:

The first and arguably most significant risk is that of messages being lost due to users deleting them themselves. Particularly concerning in this regard are the auto-destruct features offered by many instant messaging services, which can be especially appealing to politicians, for whom confidentiality and reputation are paramount. Additionally, mobile phones and smartphones have a limited lifespan and are easily lost or stolen. As a result, information stored on these devices can be lost unexpectedly, and backups may not always effectively safeguard against this. Only if the information is also stored in the cloud, either within the service itself or with a third party like Google Drive or iCloud, can it be recovered on a new phone.

Thirdly, relying on commercial instant messaging services, especially those storing messages in the cloud rather than locally, can leave users vulnerable to unanticipated decisions by the service providers, private companies providing them, who lack democratic obligations.13 Moreover, as most of these services are not paid for directly (at least not in currency), their contractual obligations are limited as well. For instance, when Reddit altered its chat infrastructure in July 2023, numerous users lost their messages because they overlooked the warning issued just 8 days prior, indicating that these messages would be erased.14 [7] Such risks should obviously be avoided when dealing with archives of significant value.

Strategies:

While it may have been considered in the past, one approach would be to prohibit all phones from the policy-making stage. However, today, this strategy seems exceedingly difficult and, arguably, counterproductive. Messaging services have become an irreversible component of today’s political communications landscape and considering that politicians frequently navigate between their professional and private spheres, it is unrealistic to expect them to entirely forego text or messaging-based communication. Attempting to do so would only lead to frustration and skepticism.

A more pragmatic approach would entail implementing specific restrictions on the use of the medium. The guidelines established by and for Dutch public administrations, which prohibit the use of instant messaging services for 1) confidential and classified information, 2) personally sensitive information, and 3) administrative decision-making, could serve as a source of inspiration.15 However, it may be necessary to adapt these guidelines for cabinet communications to ensure their feasibility. Additionally, one could argue that personal and sensitive information exchanged between political key figures might hold historical significance, even if they are not related to concrete decisions.

Not so much an alternative as an additional strategy is mandatory archiving. However, archives do not entirely control this strategy, as it is determined by law. In Belgium, for instance, the archives of ministerial cabinets are legally considered private archives and therefore not subject to the supervision of the State Archives. Ministers and their policy cells—which are particularly large and influential in Belgium16—can decide autonomously on the management of their archives. This has commonly led to questionable information management practices, such as ministers saving their digital archives to a USB stick for private use and then having their devices wiped. [14] But even in countries where ministers and their staff are supervised by the national archive, it is almost impossible to prevent messages from being lost or destroyed. After all, only the conversational partners know which conversations have taken place.

Equally crucial, therefore, regardless of the legal situation, is to raise awareness through clear communication and guidelines. This strategy has already yielded some results in Belgium. A spokesperson for the prime minister informed Lars Bové, investigating for De Tijd and EU Observer, that “Based on the recommendations made to members of the government, ministers should be careful when, for example, deleting text messages that could nevertheless have significant historical value.” [9] The end of the legislative term will reveal the extent to which this awareness has been put into practice, particularly regarding texting and instant messaging.

Going beyond mere guidelines, a practical and relatively straightforward strategy would be to ask/require ministers, and ideally their advisors as well, to sign a declaration at the beginning of their tenure. In this statement, they affirm their understanding of the guidelines and pledge to adhere to them. Such a measure could also effectively enhance their awareness regarding the blending of personal and professional communications.

A final crucial element in preventing the loss of messages from political officeholders is to provide practical instructions for archiving instant messages, especially when manual archiving is necessary (cf. infra). These instructions should be clear, easy to apply, time-efficient, and regularly updated to keep pace with the evolving landscape of applications and their functionalities. Several countries, including the Netherlands and Norway, have already implemented this strategy.

Appraising messages

The challenge:

The medium presents several challenges in terms of appraisal, notably the highly individual nature of the devices, which are often used for both personal and professional communications. Another difficulty arises from the rather unstructured content, which can typically only be classified and searched based on full-text and the rather random division and names of chat groups.

Strategies:

One strategy is to let ministers themselves select messages based on their intuition regarding their historical value. While this approach is highly unattractive to archivists as it contradicts fundamental principles of archival practice, it may be the only feasible option when manual capture is the only option. This strategy also has the benefit of being least concerning to the archive producer. In May 2023, one of the Belgian vice-premiers stated that they had no issue with archiving messages, under two implicit conditions: when the content of the messages is related to the minister’s mission (“quand les messages sont liés à la mission du ministre”) and when the minister retains control over the selection process (“pour autant que le ministre conserve son pouvoir d’appréciation”). A similar stance was echoed by the prime minister's office. [2]

A refined version of this strategy involves allowing the archive producer to undertake the selection while providing clear criteria. In this scenario, the directives should be explicit and well-founded to ensure they are embraced and adhered to by the target group. As most archives already have criteria for e-mail management, these could be, and it would be beneficial for the sake of coherence if they were extended to encompass mobile messages. For instance, the State Archives of Belgium advise archiving emails if they contribute substantively to a file, contain essential information for colleagues to follow up on a file, or involve decisions relevant to the organization or legal commitments. Extending these principles to mobile messages seems only logical, and thus forms our current starting point for advice.

A third strategy, especially beneficial for members of government, given the importance of their office, is preserving their entire mailbox. The so-called Capstone approach, initially applied to email, was extended to instant messages by the US' National Archives and Records Administration in early 2023 [10] and adopted soon after also by the Nationaal Archief of the Netherlands. [11] This method requires minimal effort from archive producers. However, it comes with the challenge of potentially containing a significant amount of personal information. To address this concern, assurances should be provided to the individuals involved regarding future selection processes or strict rules should be established for accessing and consulting these messages, both by researchers and archive staff.

A fourth strategy for selecting messages involves appraising messages group by group, recognizing that certain conversations with specific individuals may contain more important or historically valuable content than others. The State Archives or recognized private archive institutions could determine which conversations between federal ministers and their employees have historical value. This selection policy is nearly as straightforward to implement from a technical standpoint as the Capstone approach, and it enables a degree of consideration regarding personal versus professional conversations. A significant downside is the need to follow up on what groups exist, which is only possible when the archive works in close collaboration with the archive producers.

For any of these strategies to be effective, it is crucial—though not yet the case in Belgium, as previously discussed—for government officials and their staff to separate their professional communication onto two different devices or at least two different SIM cards. Without this distinction, it is practically impossible to differentiate professional messages from private ones, potentially providing individuals with an excuse to avoid archiving and/or disclosing messages.

Another challenge with any strategy is obtaining general consent for archiving from all members, especially in group chat scenarios, particularly when not all participants are legally required to archive their messages. For instance, Jeroen Buysse, a digital archivist at Liberas, the Belgian Liberal Archive, mentions the existence of a WhatsApp group of the ministers from the Liberal party, which would be interesting to archive, but to which he doesn’t expect to ever receive access: "There is even a procedure to export and store those WhatsApp conversations. But ministers obviously don't want that, a matter of professional secrets and confidentiality. [Er bestaat zelfs een procedure om die Whatsappgesprekken te exporteren en op te slaan. Maar dat willen ministers vanzelfsprekend niet, een kwestie van beroepsgeheimen en vertrouwelijkheid.]” [12]

Capturing messages

The challenge:

In contrast to documents and data sent from business laptops over controlled networks and stored within official information systems, text messages and instant messages sent from mobile devices do not pass through any official server, evading control and archiving mechanisms. To ensure their preservation, action must be taken by the sender or recipient(s). Failure to do so results in the automatic loss of these messages. An illustrative case is a request for information under the Dutch Freedom of Information Act in the Netherlands from 2019, which could not be fully accommodated due to a significant portion of the phones used during the relevant period being deprecated and replaced in the meantime. (“Het Wob-verzoek gaat over de periode van 1 oktober 2015 tot en met 9 maart 2016. Inmiddels is een aanmerkelijk deel van de in deze periode gebruikte telefoons afgeschreven en vervangen. De gegevens die op deze telefoons waren opgeslagen kunnen nu niet meer worden achterhaald.”) [13]

Strategies:

One straightforward method to archive messages, irrespective of the application or technical resources available, is by taking screenshots.17 This approach is not without its advantages: messages can be archived selectively where a comprehensive approach is not necessary or desirable. Moreover, the resulting images are relatively difficult to manipulate compared to the text or text files that result from other manual export methods, thereby enhancing their trustworthiness as records. Unfortunately, screenshots are large in size, contain limited metadata, are not immediately searchable, and, above all, are very labor-intensive to create, except in very occasional and piecemeal instances. This is particularly problematic for busy people such as politicians and other high office holders. Images will also be heavily compressed as a consequence of this method, and other media files sent as attachments will not be archived at all.

To save multiple messages at once, one can use the built-in export functions, if the application offers them. For instance, WhatsApp allows users to export conversations to a TXT file. Alternatively, users can choose to preserve media such as photos, GIFs, videos, and sound clips by exporting to a ZIP file containing a TXT file and these media files. This approach is much more feasible, and the resulting files are more user- and preservation-friendly. However, there are drawbacks to consider as well.

First and foremost, authenticity can be an issue, as the person exporting the messages could potentially manipulate them unnoticed before they enter a controlled environment such as a document management system. It is optimistic to expect that archives will obtain multiple copies of the same conversation from different people for verification. Additionally, only visible metadata are retained, including the date, time, and sender of the message. The sender's name appears as saved in the user's phone book, or only the telephone number if the sender is not saved as a contact. Message reactions, shown in a separate box at the bottom right of a message, are lost as well, despite their potential significance.18 Finally, exports may be less complete than they appear. WhatsApp limits exports to 10,000 messages including media or 40,000 messages excluding media, potentially covering only the last few months of intensive conversations. Thus, users are still required to go through the necessary steps regularly to obtain a full picture, or to entrust their device to a third party.

The backup feature falls short as a solution to this problem. While users in most applications can easily create a complete copy of their conversations with a single button press, these backups are primarily intended for transferring data between instances of the same application on different devices. Consequently, the file structures tend to be complex, and the file formats are often specific to the application. Choosing this preservation strategy would necessitate intricate processing at a later stage or emulation techniques to interpret the data effectively.

In other words, among all manual methods, none seem truly user-friendly, with screenshots—the method offering the most guarantees in terms of authenticity—being the least user-friendly of all. To ensure the integrity of conversations and maintain a comprehensive record, achieving some level of automation seems imperative, especially for the target group at hand.

Technical solutions for (semi-)automatic archiving of texts and instant messaging do exist and in fact have been available for nearly as long as texts and instant messages themselves.19 Tailored to assist businesses with risk management and compliance, these solutions typically direct messages to a controlled environment, such as a mailbox. Alternatively or additionally, they offer a desktop application that allows users to search, organize, and select messages from various communication channels, including texts and instant messages, as well as email and social media. The messages are captured either by using a client that mirrors the messaging application, a wrapped version of the application, a middleware retrieving messages in real-time or a dedicated messaging application.20 [14]

While they significantly reduce the effort required by archive producers to store their messages, these solutions do require some action(s) to initiate the archiving process—whether it's scanning a QR code to link the phone to the archiving service, switching to a different phone number associated with the archiving service, or installing and using a new service or a modified version of a familiar service. Although these steps are relatively minor, they can pose significant barriers for political officeholders, especially if they are reluctant to allow the archiving of their messages in the first place. There is also a risk that these altered or alternative platforms are less versatile than the platforms people are accustomed to, which may discourage their use.

Moreover, countries subject to GDPR privacy rules and/or national security requirements, cannot always use these archiving services, most of which are headquartered in the US.21 In such cases, governments may need to partner with a company based in their country or region, or at least one capable of storing the data there. Even more favorable, perhaps, is the development of an in-house archiving solution. This approach avoids vendor lock-in, ensures complete control over all features, and allows full oversight of the servers used for storing and transmitting messages. It is this final argument—security rather than archiving—that has driven some governments to develop their own chat applications.22

Transferring messages to the archive

The challenge:

In addition to establishing a secure transfer protocol, the primary challenge is persuading the target group to transfer all historically relevant texts and instant messages to the archive, even in the face of legal obligations. While archival institutions already encounter difficulties convincing key political (and administrative) figures to transfer their emails, convincing them to entrust archives with their mobile messages is even more challenging.23

Strategies:

Once messages are stored as files on the phone of the archive producer or another device, they can be transferred using the same procedures typically employed for other types of archives, or through any secure transfer protocol. Ideally, the process should be both secure and simple, requiring minimal effort, perhaps no more than the push of a button, enabling even busy officeholders to initiate a transfer without the need for a trusted intermediary.

Alternatively, a professional from the archive can personally retrieve the messages from the cabinet, facilitating an offline transfer method without the involvement of any additional trusted individuals. If needed, the archivist could assist the political officeholder with the capture process as well.

Another approach is to collect all professional phones at the conclusion of the legislative term and transfer them to the archives. While this strategy is the least time-consuming for the archive producer, it demands a high level of trust from the cabinet towards the archives—or may require a strict legal obligation—and does not prevent messages from being deleted. To implement this approach effectively, all individuals involved must also possess a professional phone and be fully informed that the device will eventually be transferred to the archive.

Finally, an automated archiving solution presents the possibility of establishing a direct link with the repository which serves as the final destination for the messages, although it is unlikely that individuals will permit a transmission without any prior selection, as indicated earlier.

Irrespective of the chosen methodology, it is imperative that the messages are transferred alongside all requisite metadata to enable the reconstruction of the contextual framework in which they were exchanged. This comprehensive approach is indispensable to ensuring their utility as authentic historical artifacts and safeguarding against potential misinterpretation or misuse.

Some thoughts on preserving messages

In the preceding paragraphs, we explored the intricacies of capturing and obtaining messages from archive producers. Beyond this initial step, archives must address the task of sustainably preserving these messages and ensuring their ongoing accessibility. Fortunately, these are challenges archives are familiar with and often have already solved in part or in whole. The same procedures and tools that are used for redacting private information from emails may, for instance, be used for texts and instant messages as well.

It is to be expected that best practices and specifications will emerge for storing mobile messages. What procedures archives will have to apply to create compliant archival information packages depends on the format(s) in which messages were captured. For sceenshots, this will probably involve the application of optical text recognition and extraction of the data and metadata. Tools or AI models for this will certainly be welcomed with gratitude by many archivists.

The way the messages are displayed should reflect the type of medium from which they originate as part of the context, we believe. A more difficult question is when and under what circumstances they can be made accessible to the public. We propose exchanging views with archivists from different countries to reach a common conclusion and be able to propagate it with greater authority.

Conclusion

In the past two decades, mobile messaging has become a major communication channel in policy making. While official decisions are still conducted through formal channels and documented in standardized formats, they are often preceded by informal consultations and agreements communicated via text or instant messages. Therefore, it appears imperative now, for public confidence in government and in the fight against disinformation, corruption and other threads to democracy, to start taking this type of archive seriously and begin implementing measures to prevent important messages from ending up in the digital trashcan.

Due to their unique nature, properties, and production device, texts and instant messages present a particularly challenging category of data, fundamentally distinct from traditional documents or other types of data archives typically encounter. Unlike other data, they must be exported from the device on which they were created before they can be preserved in an information management system or filing system. Although more empirical data is needed for a more definitive assessment, the main challenges in archiving mobile communications appear to be not of a technical nature, but related to human factors, and to lie in preventing messages from being accidentally lost or purposefully erased.

Archives are inevitably dependent on the collaboration and goodwill of archive producers to develop and implement adequate solutions, so much is clear. This challenge is particularly pronounced in the Belgian context, where cabinet archives are not legally considered public archives, but is pertinent in any legal context, as supervisory bodies mostly lack a comprehensive overview of information flows via mobile channels. We can, however, enhance the likelihood of successful preservation in a number of ways, such as by raising awareness, providing guidance, support and training, while also augmenting our own experience with the selection, transfer and preservation of texts and instant messages through experiments, comparative research, and intra- and international collaboration.

Moreover, archives can work on building a trusting relationship with the archive producers. This endeavor, in turn, requires them to optimize the security of their electronic repository, such that depositors can be assured of the safekeeping of their sensitive content, and to stipulate clear rules for data consultation. Both in interpreting the principles of freedom of information and in establishing modalities for studying mobile messages in the long term, a careful balance must be maintained between maximum openness and sufficient confidentiality, to ensure that archive producers preserve their messages and entrust them to an archive. Determining where this balance lies is a topic we are keen to discuss with fellow archivists and stakeholders.

While the real-life effectiveness of the proposed strategies should be confirmed through further experimentation and implementation, we hope that our findings thus far, along with the considerations outlined for each method, will prove valuable to other archives in their search for an adequate, sufficient, and practicable strategy for archiving texts and instant messages by political officeholders, and by extension, other archive producers.[4][7][8][9][11][12][21][22]

ACKNOWLEDGMENTS

I want to thank Kathleen Devolder, Johan Van Der Eycken, Oliver Lenaerts and Karel Velle for their useful suggestions for this paper and for support.