This article summarises preliminary reflections for mediators on social media codes of conduct. Such codes of conduct, to be agreed upon by conflict stakeholders, would aim to prevent and mitigate the use of social media to exacerbate conflicts or jeopardise peace processes. Such codes of conduct would apply to both social media content (specific types of posts and comments such as hate speech) and social media behaviour (specific uses of social media such as coordinated efforts to manipulate public debate).
Social media can be used to exacerbate conflicts or widen political divides through, among other things, the dissemination of hate speech and disinformation, the identification and targeting of individuals or constituencies, political manipulation through targeted campaigns, the coordination of violent actions, or the recruitment of combatants.
Social media can also be used to jeopardise peace processes by, among other things, spreading leaks of confidential information about peace negotiations or agreements including the location of talks, the identity of negotiators, and the positions of participants. In addition, social media can be used to target participants and mediators, their families and the constituencies they represent, or to contest the legitimacy of the process or their participants.
Prior to the drafting of a social media code of conduct, it is essential to conduct a thorough analysis of the exact context such an agreement would be developed in. This includes identifying social media platforms used, main conflict stakeholders and influencers present on these platforms, and problematic behaviours and content exacerbating conflict or negatively impacting ongoing dialogue efforts. Without this analysis, a code of conduct risks being too vague or inadequate.
Limitations: a code of conduct is only relevant in contexts where social media is proven to have a negative effect on conflict dynamics or risk jeopardising a peace process, and where existing norms are insufficient to prevent or mitigate such uses. A code of conduct is likely to be rejected if it is perceived as unnecessarily impeding fundamental freedoms.
Codes of conduct can be signed by anyone, including representatives of conflict parties (governments and armed groups), political parties, civil society organisations, diaspora, and private sector actors. To be successful, signatories should include most, if not all, individuals and groups who can meaningfully contribute to the implementation of the code of conduct and whose actions could exacerbate conflicts or jeopardise the successful proceedings of a peace process. Ideally, signatories should adhere to the code of conduct voluntarily.
Limitations: because of fragmentation, dissent or ignorance of the law, signatories might not be able to control the constituencies they claim to represent. This challenge is not exclusive to the social media realm but should nonetheless be taken into account in the design of the specific clauses, including calls for public condemnation of violations and the proactive encouragement of positive online behaviours.
3. Potential clauses
Clauses to be included in a social media code of conduct need to address the particular challenges of the society the code is being developed in. The following types of clauses could be included:
ο Refrain from contributing to the dissemination of problematic content that either exacerbates conflict (provocation, intimidation, threats, insults, hate speech, polarising language, disinformation, etc.) or undermines dialogue efforts (breaches of confidentiality, direct attacks on the dialogue process or actors involved, leaks, etc.).
ο Refrain from engaging in problematic behaviours exacerbating conflict or undermining dialogue efforts (conducting information operations, using fake accounts to misrepresent political actors, microtargeting along conflict lines, recruiting combatants on social media, etc.).
ο Actively promote specific content (such as publicly countering misinformation and hate speech, promoting content that supports dialogue efforts, etc.).
ο Actively engage in specific behaviour (such as educating and informing the public about the code of conduct, promoting digital literacy, working to combat polarisation, etc.).
Limitations: a code of conduct can only be effective if it clearly defines what constitutes a violation and empowers a monitoring body with sufficient means to address such violations. If definitions are too broad or listed violations too numerous, a code of conduct is unlikely to be implemented.
To ensure compliance, a code of conduct requires:
In addition, legislators might play an important role in adopting laws which reflect the terms of the code of conduct and thus strengthen the legitimacy of the agreement and the likelihood of it being enforced.
Limitations: content and behaviour on social media can be difficult to attribute; private spaces on social media might be difficult to monitor; and codes of conduct can threaten free speech and be used for censorship. No social media code of conduct can be expected to be fully respected. Expectation management is therefore key for the success of such an agreement.
5. Existing experiences with social media clauses
While social media codes of conduct to prevent and mitigate the use of social media to exacerbate conflict and jeopardise peace processes have, to date, never been signed, some conflict stakeholders have in the past agreed, both informally or in writing, to abide by specific behaviours on social media.
For example, David Lanz and Ahmed Eleiba note that:
During the final stage of the Iran talks, the negotiators agreed to refrain from communicating via social media. They notably left it to the EU High Representative Federica Mogherini to announce on Twitter the successful conclusion of the Joint Comprehensive Plan of Action in July 2015. 
In addition, clauses referring to social media have been included in agreements signed in the past five years. For example, in Myanmar in June 2015, 68 political parties signed a code of conduct for the 2015 general elections that included the following section:
The Parties commit themselves, when addressing the public at political rallies or as part of their communications through mass media, including the social media, to refrain from:
a. any form of intimidation or incitement to violence vis-à-vis any person or group of persons or beliefs;
b. defamation and incitement to hatred, or accusation of apostasy, treason, terrorism or any other similar serious charges;
c. fuelling regionalist, racial, sectarian or tribal trends that could threaten national unity;
d. insult, libel and degradation.
Similarly, other agreements, including in Syria, South Sudan and Libya, have included clauses where, for example, signatories agreed to “a cease of provocations on social media such as Twitter and [similar platforms]”, “not engage in any form of hostile propaganda or hate speech, or use any media, including social media to foment ethnic or sectarian hatred” or “reject the pages on social media sites that call for fighting and sedition.” 
Finally, in November 2019, a group of leading Libyan journalists, bloggers and social media influencers attended a workshop on “Combating Hate Speech and Ethics of Journalism”, held in Cairo, after which they released the following recommendations for social media users, activists and media outlets in Libya:
Social media is an integral part of today’s conflict landscape and mediators have no choice but to address its impact on conflicts and peace processes. The idea of a social media code of conduct is not new  and this article only constitutes preliminary reflections on a topic which merits further in-depth research, consultations and testing. More comprehensive guidelines could be established through the incorporation of lessons learned during the development, implementation and monitoring of electoral codes of conduct or ceasefire agreements. Peacemakers would also benefit from maintaining regular contact with social media companies, who have extensive experience with monitoring online content and behaviour. The topic would benefit from further research into how social media clauses included in past agreements were developed and dealt with and how bodies created to monitor agreements, which included social media clauses, have approached violations. It will also be important to work with stakeholders from ongoing or latent conflicts to explore how best to include social media clauses in agreements or to develop social media codes of conduct. While the ideas shared in this article are non-definitive, we hope that they can contribute to needed reflection, and help further drive conversations on the topic.
 Jonathan Harlander is a Project Officer at the Centre for Humanitarian Dialogue and the incoming Senior Manager, Peacetech at Crisis Management Initiative. This article is based on a presentation the author delivered in February 2020 at the Data for Peace and Security Conference, in The Hague. The author would like to thank The Netherlands Ministry of Foreign Affairs for this opportunity. The author would also like to thank all the members of the CyberMediation Network for providing a space for reflections and exchanges on the topic of digital technologies and mediation, and Dr Katia Papagianni for her feedback and support throughout the development of this article.
 Maude Morrison is Deputy Director at Build Up. The article reflects work done during a partnership between Build Up and the Centre for Humanitarian Dialogue. A particular contribution was made by Build Up’s Director, Helena Puig Larrauri, who provided comments and drove much of our thinking on this topic throughout the partnership.
 David Lanz & Ahmed Eleiba, The Good, the Bad and the Ugly: Social Media and Peace Mediation, swisspeace, Policy Brief 12/2018.
 Code of Conduct for political parties and candidates, 26 June 2015.
 Agreement on the Cessation of Hostilities, Protection of Civilians and Humanitarian Access, Republic of South Sudan, 21 December 2017.
 Ceasefire between the City of Tarhuna and the City of Tripoli, 21 September 2018.
 United Nations Support Mission in Libya, Libyan journalists, activists and social media influencers meet in Cairo and agree to combat hate speech, Cairo/Tripoli, 13 November 2019.
 “Mediators might need to consider working with the parties to establish ground rules or a code of conduct on the use of social media as well as on what information relevant to the process should be shared with the broader public and when.” In United Nations Department of Political and Peacebuilding Affairs and Centre for Humanitarian Dialogue, Digital technologies and mediation in armed conflict, March 2019, p.29.