The Cross-Platform Evasion Toolbox of Islamic State Supporters

Ayad offers these platform recommendations to stem exploitation:

The Islamic State supporter platform exploitation toolbox poses a serious challenge to the health of the overall information ecosystem online in the Middle East, Africa and beyond. Narratives espoused by Islamic State supporters undermine governments, technology companies and wider civil society by suggesting they are puppets of Western states and incompatible with the Islamic State’s ideology. They are often coupled with sectarian tropes focused on the real and perceived injustices faced by Muslims at the hands of various governments and their security forces. Their capacity to camouflage as news outlets requires a deeper understanding of Islamic State networks online and how they mobilise to survive in the face of increased moderation. Practical steps are needed to deal with these challenges:     

First, there is a clear need for more expert-led moderation and investigative tracking of accounts and their spheres of influence. This should come with greater investments in Trust and Safety divisions in smaller or less profitable markets. The ability of these support networks to survive takedowns necessitates in-depth tracking on all platforms in question and the use of subject-matter, linguistic and contextual expertise in several regions, such as East, Central, and West Africa. 

A model of cross-platform network coordination around content removal should be developed for these networks rather than relying on piecemeal takedowns that focus on single-actor behaviours or single platforms. Technology companies can demonstrate the same coordination as Islamic State supporters through entities such as the Global Internet Forum to Counter Terrorism (GIFCT). Equally, Europol demonstrated this possibility in 2019 when it targeted Islamic State channels in a wide-ranging, multiplatform takedown effort.

Issues persist regarding the role of artificial intelligence in moderation efforts. To enhance the capabilities of artificial intelligence to recognise content that has been ‘scribbled’, systems should be trained on ‘scribbled’ content that has been previously flagged and identified via a manual review. Further, creating mechanisms by which companies can detect variations in language that allow supporters to continue to publish inflammatory content is an important endeavour.

While platforms such as Facebook and X continue to play pivotal roles in assisting independent media outlets globally, Islamic State news outlets are disseminating terrorist content and disinformation under the guise of independent and ‘objective’ news sources. Unlike independent journalists, outlets, and civil society groups, however, these outlets are only focused on Islamic State actions and objectives. 

Taken together, these platform evasion and exploitation tactics and strategies by the Islamic State seem to indicate that the current moderation strategies of platforms are ill-equipped to deal with a much more nimble-footed set of users adapting to platform moderation techniques. What is needed is a cross-platform acknowledgement of these strategies and an adjustment in how moderating these Islamic State support communities can inhibit further support in an era where they evade and exploit platform lapses to remain and expand.