Click here to read our latest report “Transmisogyny, Colonialism and Online Anti‐Trans Activism Following Violent Extremist Attacks in the US and EU”

Falling Through the Cracks: Gender Bias in Online Counter-Terrorism and Prevention

Falling Through the Cracks: Gender Bias in Online Counter-Terrorism and Prevention
3rd July 2023 Bàrbara Molas
In Insights


In Orientalism (1978), the book that shaped post-colonial studies, Edward Said tried to understand how and why colonised countries developed the way they did in a postcolonial era. Inspired by post-structuralist Michel Foucault, his book showed that academia was closely connected to the colonial system, as it was the Western production of knowledge that defined power and the way individuals are organised around it. With his counter-hegemonic thinking, Said’s goal was to have academics admit that they had played a key role in sustaining the prejudices that allowed colonialism to maintain itself. While Orientalism did not address how “imperialism depends … on sexuality and gender for its viability”, it did constitute a turning point by which all disciplines in the Humanities and Social Sciences were encouraged to question the dominant epistemologies informing their research.

In this Insight, I attempt to apply the same critical approach to the field of counter-terrorism (CT) and prevention of violent extremism (P/CVE), by interrogating how we monitor online threats. With a focus on misogyny, I argue that our understanding of the threat landscape relies on research that has defined ‘extremism’ based on reactions of the criminal justice system towards men. I argue that, online and offline, the threat of female extremists is downplayed and dismissed based on the assumption that violent behaviour in women is ‘gender atypical’. This is an example of misogyny, a form of sexism that perpetuates stereotypes and prejudices towards women, removing their agency and helping maintain unjust cultural customs and systems. The consequence: women fall through the cracks of our CT and P/CVE efforts.


According to Omi Hodwitz, across multiple offence types, female offenders are treated more leniently than males. In this context, violence is considered gender atypical, which leads to a decrease in leniency for violent offences. The adoption of gender-based assumptions when determining the severity of a sentence implies that female individuals tend to be considered an unforeseen threat. On the one hand, this encourages efforts in online monitoring to focus on male subjects, or in male-dominated circles. On the other, since violence in counter-terrorism is considered to be the ultimate expression of extremist and terrorist behaviour, such an assumption around gender atypical activity implies that women are by default treated as a minor threat.

If the literature and the judicial understanding of (and production of knowledge around) extremism are based on the belief that violence emerges mostly from men, then the tendency among CT and P/CVE experts would be to understand violent extremist ideas based on how men express them. But narratives and ideologies that aim to fuel violence are endorsed and manifested differently depending on how a subject self-identifies. Here, gender definitely matters, but only to the extent that it means something to the radicalised user – not to the monitoring entity. As stated by Alexandra Phelan, “Women and men experience terrorism differently” and so divergences exist at the levels of “radicalisation to violence, terrorist motivations, roles, and factors that can sustain involvement” in each context.

This phenomenon has been explored by Cat Tebaldi, who showcased that women express their radicalised beliefs very differently from men in online far-right circles. Images of a family stamping sugar cookies with “intricate runic patterns”, or canning vegetable jards “surrounded by laughing blonde children”, should not be seen as a threat per se, but when such images are accompanied by calls for “reviving folk vitality”, then we should be concerned. This is what Tebaldi calls “digital traditionalism”, a female-led form of online radicalisation that thrives on social media. This is thanks to the algorithmic presupposition that women disseminating anti-feminist lifestyles is not a manifestation of, or a trigger for, violent extremism. 

Similarly to Tebaldi, Eviane Leidig has demonstrated that on mainstream social media platforms like Instagram, radicalised couples are able to survive due to only one of their accounts being banned. The way that a male partner manifests his radicalisation activates the flagging system, while the woman is able to keep spreading harmful content, circumventing Terms of Service using the stereotypes embedded in our own monitoring and reporting systems to their advantage.


Our biased monitoring practices in the flagging process show that we don’t recognise that the threat landscape is shaped by various forms of engagement, including gendered interactions and behaviour. This is because our responses to terrorism are significantly influenced by the ‘sex’ of our suspect. This includes the way women in online extremist circles express their beliefs. In order to counter extremist content, we need to learn more in order to detect and make evidence-based risk assessments on how much their harmful activities could act as action triggers leading to offline violence. Most prosecuted individuals for terrorist offences are male. However, studies show that this is not necessarily due to men being responsible for most offences, but because the evidence is androcentric.

Evidence associated with women’s involvement in terrorist activities has remained on the periphery of counter-terrorism discourses, “either viewing them solely as victims or as … incidental associates on the primary terrorist actors”, as put by the United Nations. Indeed, we still struggle to see that involvement in terrorism does not require active violence, but it should be punishable when support exists (or existed) and can be proven in other ways. For example, women in the so-called Islamic State (ISIL/Daesh) assist the system by providing for their families, and ensuring the survival of their male spouses in their quest against ‘enemies of the state’. But, because resources to gather this type of evidence are fewer than those invested in locating and prosecuting men, criminal processes tend to favour women, who are accomplices in terrorist activities. This same problem is reflected in the online space where flagging systems are designed to identify harmful content on the basis of male-gendered extremism. This helps to build a limited knowledge database of evidence of online threats, thus informing a faulty CT and P/CVE practice.

There are distinct ways in which women and men are involved in and affected by terrorism, but we are not collecting enough data to prove it. This is due in part to multiple biases that determine our research questions and focus, and also to the existing gap between groundbreaking academic work on the subject and the work of practitioners on the ground. Even though academia has made important progress in better understanding gender dynamics in terrorism, the implementation of such progress has been slow. On the one hand, “institutionalised assumptions about the role of women in political violence”, even when acknowledged, still constitute an impediment to change. On the other hand, the assumption that online and offline radicalisation is a ‘gender-neutral’ process is still widespread in the safety and security sectors.


Gender biases perpetuate the field, making our job less effective. When we ignore the participation of women in terrorist activities, we are missing half the picture.​ Firstly, our lack of understanding​ of gender dynamics in the threat landscape implies that we only partially comprehend how terrorist and violent extremist organisations use gender to shape their own agendas, recruit, and implement policies that can espouse a distinct, gender ideology. This is evident in monitoring practices that dismiss gendered forms of engagement, allowing bad female actors to continue disseminating hate speech and facilitating radicalisation (affecting all genders) across the online space. 

Secondly, in dismissing part of the threat on the basis of gender prejudice, we are also making gender-based threats less evident to our staff. If we inform our practices based on the idea that women are expected to express themselves in ‘gender-typical’ ways, how can mainstream misogyny be seen as a threat? Finally, if we prosecute more men than women on the basis of assumptions around the latter’s limited agency, we are robbing radicalised or at-risk women of the opportunities to participate in deradicalisation, rehabilitation, and reintegration programs.​ Research by the UN suggests that women “tend to receive more limited rehabilitation and reintegration support, thus putting them at potentially greater risk of recidivism and re-radicalization and potentially undermining their successful reintegration into society”.

Countering gender bias in our CT and P/CVE practice needs to start by raising awareness of the different ways in which men and women are affected by terrorism and by government responses to terrorism. This includes acknowledging that radicalisation processes are gendered and that women in extremist communities pose a real threat. But acknowledgement can only take place after ensuring the systematic application of a gender lens in our practice, which requires increasing the participation of women in the field. This will facilitate building knowledge on and safeguarding the evidence that showcases the role that gender plays in shaping terrorism and violent extremism. Supporting such change will allow for an improved and more nuanced capacity to identify threats, and inform more effective early warning signalling mechanisms – isn’t this what prevention of violent extremism is all about?