1 The Founding of Syrian Archive and Mnemonic

Mnemonic’s origin is Syrian Archive (n.a.). Syrian Archive was founded in 2014, but it has its roots in the Arab Spring. In 2011, our founder Hadi al Khatib and a few friends realized that although there was a lot of content being posted on Facebook from Syria, there was no news published in English and no summary of the videos that were being published on a daily basis. They started doing so in March of 2011 on their page the Syrian Uprising Information Center (Syrian Uprising, n.a.).

In 2013, Hadi left Syria to pursue this work from outside the country, but continued to collaborate closely with videographers. He started to hear the same problems over and over again — cameras and phones lost with masses of photos, and irreplicable Facebook pages and accounts suspended. It’s important to keep in mind that at this point, Facebook had created very few support systems for human rights defenders, and the Syrian government was torturing people for their Facebook passwords, something that human rights defenders have told NGOs is still happening (Freedom House, 2020).

So, in 2014, Hadi founded Syrian Archive, with the goal of preserving as much digital memory of the Syrian conflict as possible - for potential use as evidence in legal proceedings of human rights violations and fact-finding missions, but also to ensure that people’s experiences were not forgotten. Those dual uses of this content are equally important, though the focus is largely on the potential use of this content as evidence and for fact finding.

2014 was a very different time from today. The importance of this content had not yet been recognized. There was no one doing what Syrian Archive did, although there were others in our community who were collecting and using social media content, such as the Syria Justice and Accountability Center and the Violations Documentation Center (which shifted into the Syria Center for Media and Freedom of Expression) (Syria Justice and Accountability Centre, n.a.; Syrian Center for Media and Freedom of Expression Homepage, 2020). Funding was hard to come by, and standards were non-existent. Systems and processes had to be built from scratch.

Hadi worked with his colleague Jeff Deutch and others for the next several years to develop a standardized workflow and the technical tools needed to preserve social media content at scale in the most usable formats as possible (Mnemonic, n.a.). We started to create meticulously verified and curated investigations into specific incidents and types of attacks, such as our databases of attacks on medical facilities and our incident report into the 2017 bombing of a mosque in Aleppo by the United States (Syrian Archive, 2020). We learned from the verification work that groups like the open source investigation collective Bellingcat were already doing similar work, as well as others utilizing open source investigative methodologies like the Organized Crime and Corruption Reporting Project (Organized Crime and Corruption Reporting Project, 2007; Bellingcat, 2022). The open source investigations and evidence community was even smaller then than it is now, and so many people learned from each other through working together, including Syrian Archive, Amnesty Internationa’s Digital Verification Corps, and Bellingcat. We also learned from the standard operating procedures of international justice digital forensics experts. While these tools and processes were not developed for mass archiving or large-scale analysis, we adapted them to fit our purpose. We were more and more often providing support to other organizations, and even founding new archival projects like the Yemeni Archive (Yemeni Archive, n.a.). We also found that our experience with the impacts of content moderation added an important element to increasingly public discussions about platforms’ tools and policies.

In 2017, as we worked with myriad partners and advocated against the incredibly rapid removal of Syrian content from YouTube, the idea of Mnemonic as an umbrella organization developed. Syrian content started to disappear from YouTube at an incredibly high rate immediately after the company announced it was using machine learning to detect and remove so-called terrorist and violent extremist content (El Deeb, 2017; Walker, 2017). Content that was posted from the region, or in Arabic, was caught up and improperly deleted (Al Jaloud et al., 2019) That year saw another important milestone as social media content received high-level evidentiary recognition — for the first time ever, this content provided the basis for an International Criminal Court warrant, which was issued against senior Libyan military commander, Mahmoud Al-Werfalli (Mierke, 2021).

2 Mnemonic

Mnemonic now exists as the umbrella organization for all of our work. We officially launched Yemeni Archive in 2018, and with our partners Gisa we launched the Sudanese Archive in 2019 (Sudanese Archive, n.a.). At the beginning of this year we launched the Ukrainian Archive (Alfred Landecker Foundation, 2022). Mnemonic’s structures provide support to each of those archives, and we provide rapid response, long-term capacity building, thought leadership and policy advocacy. Mnemonic’s work in specific countries or conflicts is often not public. We choose where to work based on many factors, but one essential point is whether there are local people who reach out to us or who want to work with us. We do not work on archiving without local knowledge.

The archives that are officially under Mnemonic’s umbrella each have their own websites, with the exception of Ukrainian Archive because it is brand new. These websites publicly publish only a small amount of archived content, including content linked to specific investigations and incident reports or as part of larger thematic datasets, ensuring that these memories are not erased. Mnemonic also handles requests for specific pieces of content from a variety of sources, including justice mechanisms, which we carefully consider.

Our policy advocacy largely grew out of that 2017 mass deletion of Syrian content (Khatib & Kayyali, 2019). Hadi and Jeff started engaging regularly with social media companies on content takedowns, pushing for companies to restore removed human rights documentation and to improve their tools and policies. We realized that we had an incredible opportunity to track takedowns by comparing our own collections to what was still online, resulting in the creation of our Lost & Found project (Syrian Archive, n.a.). Our learnings contributed to advocacy against the EU’s terrorist content online regulation (Kayyali, 2022a). The advocacy has grown since, with Mnemonic bringing real-world experience to often abstract content moderation discussions.

Mnemonic also conducts rapid response work. Activists come to us to rapidly archive content before social media platforms can take it down. This work sometimes overlaps with our trainings — short trainings or, when possible, in-depth trainings that can help activists start their own archives either on their own or in partnership with us.

3 Ukraine and Trends

When the re-escalation of hostilities in Ukraine started on February 24, we were already aware of an open source investigative community existing in Ukraine, which had grown out of the 2014 Euromaidan movement and the crackdowns it faced. We knew that lessons learned over the last decade through preserving and making use of open-source documentation from elsewhere like Syria could be applied to content being produced and uploaded within the Ukrainian context. Our initial rapid response support to preserve content quite quickly led to the founding of the Ukrainian-led Ukrainian Archive. We were able to move so quickly due in large part to the adaptability of the existing technical workflows and stable infrastructures we had already built to work with large-scale datasets of digital content.

Though all armed conflicts, protests, or resistance movements in which Mnemonic preserves and makes use of open source documentation differ from each other, Ukraine has been quite different from any that came before. Unfortunately, one of those reasons is that those impacted by human rights abuses are, for the most part, white and Christian. Much has been written on the disparity between responses to Syria (and other Muslim-majority countries) and Ukraine that does not need to be repeated here (Bayoumi, 2022; Rozzelle, 2022; Wamsley, 2022). But the fact is that Western governments, and their populations, overwhelmingly expressed support for Ukraine. Similarly, Ukraine has gotten more, and more sustained, media coverage.

Another reason this conflict is different is because people had unparalleled access to technology and the Internet, and they immediately started posting as much content as they could to social media. Unlike in the Syrian context, there exists myriad platforms on which to do so, and cell phones, open source investigations, and archiving techniques have grown much more sophisticated since 2011. Governments have expressed interest in supporting open source investigations. And, while it is still too early to tell the impact, the Ukrainian prosecutor has even created a portal for people to upload their own evidence demonstrating the acceptance of digital content as evidence within traditional legal structures.

What is also similar in the Ukrainian context is the need for policy solutions to the problem of mass deletion of content documenting human rights violations from social media platforms. Perhaps because of the biased interest in and coverage of Ukraine, policymakers are more willing to seek solutions. There were calls after February for social media companies to preserve all content from Ukraine that could potentially be evidence — including from members of US Congress (House Committee on Oversight and Reform, 2022; Office of the Spokesperson of United States Department of State, 2022). While the impetus to save everything that could be evidence is understandable, those policy solutions are not completely formed yet. Preservation and access to human rights documentation is a delicate balance between the right to justice and the right to privacy and security of people whose personally identifiable information is contained within that documentation.

As we wrote in a recent whitepaper, we’re not sure exactly how the problem of deletion of human rights documentation from social media platforms should be solved, but we know what shouldn’t happen — indiscriminate preservation (Kayyali, 2022b) Preservation by social media companies or third parties should have clear limits and safeguards in place. Otherwise, it has the potential to create a trove of content that could, at the end of the day, be accessible to bad-actors, including governments, who could misuse it in myriad ways. Instead, we have been calling for open dialogues amongst varied stakeholders, more clarity about how long social media companies hold onto content that they have made publicly inaccessible and a usable tool for the International Criminal Court and United Nations to request content from platforms.

4 Where Do We Go Now?

Unfortunately, human rights around the world continue to be violated. With documentation of these crimes continuing to be uploaded to social media platforms as they occur, the need to preserve this content is not going away anytime soon. Now that the public and lawmakers are aware of its importance, there is room to both craft better policies, and to adapt our own processes to ensure that the content we preserve is both useful and usable. It’s a good time to step back and assess where things stand, and to bring together the international justice community, including the ICC and UN, with groups like Mnemonic who have been talking to social media companies for many years. We want to start from a crystal clear understanding of how this content is being used now to ensure that our policy advocacy and our archiving standards are tailored to that reality. We also want to include privacy advocates, so that we can properly assess the potential negative impacts of any solution.

Finally, regardless of any of the issues that still exist, it is important to emphasize that our archives are digital memory. They are living monuments to the lived experience of human beings. While it is, of course, rewarding to know that content we have archived is being used as evidence, it is important to note that were this content to never be used by any single investigatory mechanism or in any single courtroom there is still a need to keep doing what we do. It has been more than a decade since the Arab Spring. Much of our region is facing extreme hardship, from conflicts to famine. But underlying those difficulties is still a belief in a better world, and a passion for freedom. Some lessons we learned from the Arab Spring are how to use the Internet to influence positive change in the world, to promote justice and human rights. We honor the memories of all those who have died or been imprisoned by continuing our work and sharing the knowledge they helped to build with anyone who needs it.