By Iftekharul Bashar

The recent arrests in Malaysia confirm that radicalisation within the Bangladeshi diaspora is a significant and evolving threat. This problem stems from socioeconomic factors, homeland instability, and online recruitment. A transparent, collaborative, and multi-faceted P/CVE (Preventing and Countering Violent Extremism) approach is essential to mitigate the risk.

COMMENTARY

Malaysian authorities recently arrested 36 Bangladeshi citizens in Selangor and Johor for their involvement in a radical militant movement promoting Islamic State (IS) ideology. Home Minister Datuk Seri Saifuddin Nasution Ismail reported that these individuals were forming recruitment cells, fundraising for terrorism, and plotting to overthrow the Bangladesh government.

Notably, the network, as detailed by police, raised funds using international fund transfer services and e-wallets, directing money to the IS group in Syria and Bangladesh. This method highlights a growing trend in terrorist financing, leveraging the speed, lower costs, and often less stringent oversight of digital platforms and cross-border money movement to facilitate illicit financial flows globally.

Of those arrested, five have been charged with terrorism-related offences, 15 face deportation, and 16 remain under investigation, with the police anticipating further arrests. Malaysian authorities estimate that 100 to 150 individuals are suspected to be involved in this network, demonstrating the scale of the threat they are actively dismantling.

These arrests are a stark reminder that the threat of extremism continues to cast a long shadow, not just within Bangladesh’s national borders but also across diaspora communities. The arrest of the 36 nationals is not an isolated incident. There have been previous cases of radicalisation of Bangladeshis in Malaysia.

In May 2019, a 28-year-old Bangladeshi mechanic was arrested in Kuala Kedah; he had possessed the necessary chemicals and expertise to produce improvised explosive devices (IEDs). Earlier in January 2017, two Bangladeshi salesmen, aged 27 and 28, were arrested in Kuala Lumpur, reportedly for having ties with suspected IS militants in Bangladesh and for planning to join a terror cell in the southern Philippines.

Singapore, among other nations, has also experienced similar cases in the past, notably in 2015, 2016, and 2020, highlighting a recurring pattern of Bangladeshi diaspora members being targeted and recruited by terrorist networks, including the Islamic State.

S. Rajaratnam School of International Studies, NTU Singapore, 2025. 6p.

By Johanna Kleffmann, Swetha Ramachandran, Noah Cohen, Siobhan O'Neil, Mohammed Bukar, Francesca Batault, Kato Van Broeckhoven,

Key Findings • The labelling of banditry as “organized crime,” “unknown gunmen,” or, more lately, “terrorists” has distorted this multi-faceted and still poorly understood phenomenon. Likewise, some of the frames that have been applied to banditry – particularly the farmer/herder conflict or Hausa/Fulani tensions – do not appear to fully align with local communities’ understandings of today’s evolution of banditry. Simplistic categorization and narrow lenses for understanding banditry may contribute to inappropriate or insufficient policy and programmatic responses. MEAC’s survey unearths some of the lived experiences with bandits and sheds light on the nuances of the phenomenon and its profound impact on communities. • Community perceptions of bandit groups corroborate earlier research depicting them as comprised of largely distinct, organized groups, albeit with shifting configurations and subject to fragmentation. Bandit groups operate in highly mobile, armed, and largely forestbased units that use quick-strike attacks on motorbikes against communities. Their motivations are primarily perceived as economic/financial by victims, including the notable subsection of the sample of Fulanis who have been victimized by bandits. • For the surveyed communities, weapons are the most recognizable feature of bandit groups. This bears critical implications for the potential for escalation of violence, further proliferation of illicit arms and ammunition including in the neighbouring regions, community violence reduction efforts and future DDR programming. • Victimisation experiences differ considerably with gender, age, and location in the North West. While physical violence and killings disproportionately affect adult men, sexual violence appears to especially affect women and girls (although it likely remains underreported). Variations of victimization across states indicate the volatile and dynamic nature of overall banditry presence and violence. • Banditry violence has profound and pervasive effects on the physical safety, access to income-generating activities, education, and mobility of residents in the northwestern communities surveyed. The perceived intensity and frequency of attacks are on the rise, with one in three respondents reporting experiencing weekly attacks in recent years. Close to two-thirds of respondents have family members who have been attacked by bandits.

Findings Report 36,

Geneva, The United Nations Institute for Disarmament Research ., UNIDIR, 2024,43p.

By The International Crisis Group

What’s new? Since the Taliban takeover of Afghanistan in 2021, the Islamic State-Khorasan Province (IS-KP) has emerged as a major international security threat, orchestrating or inspiring attacks abroad following a Taliban clampdown on its home turf. Although its strikes have fallen in number in 2025, its offensive could resurge. Why does it matter? Despite the recent lull, IS-KP might reactivate commanders willing to carry out attacks abroad or coordinate with other ISIS branches to launch them. Even a small number of highprofile operations – such as the March 2024 mass shooting and arson in Moscow – can cause numerous deaths and have major international repercussions. What should be done? Coordination among security services has improved, particularly in intelligence sharing and rendition. There are strong reasons not to resort to military action, but more could be done in terms of collaborating with the Taliban and Syrian governments, redefining the global anti-ISIS coalition’s law enforcement role, and supporting Central Asian countries.

Crisis Group Asia Briefing N°183

Brussels: International Crisis Group, 2025. 28p.

By Michelle Sydes | Lorelei Hine | Angela Higginson4 | James McEwan | Laura Dugan | Lorraine Mazerolle3

Background: Criminal justice agencies are well positioned to help prevent the radicalisation of individuals and groups, stop those radicalised from engaging in violence, and reduce the likelihood of terrorist attacks. This Evidence and Gap Map (EGM) presents the existing evidence and gaps in the evaluation research. Objectives: To identify the existing evidence that considers the effectiveness of criminal justice interventions in preventing radicalisation, violent extremism and terrorism. Search Methods: We conducted a comprehensive search of the academic and grey literature to locate relevant studies for the EGM. Our search locations included the Global Policing Database (GPD), eight electronic platforms encompassing over 20 academic databases, five trial registries and over 30 government and non‐ government websites. The systematic search was carried out between 8 June 2022 and 1 August 2022. Selection Criteria: We captured criminal justice interventions published between January 2002 and December 2021 that aimed to prevent radicalisation, violent extremism, and/or terrorism. Criminal justice agencies were broadly defined to include police, courts, and corrections (both custodial and community). Eligible populations included criminal justice practitioners, places, communities or family members, victims, or individuals/groups who are radicalised or at risk of becoming radicalised. Our map includes systematic reviews, randomised controlled trials, and strong quasi‐experimental studies. We placed no limits on study outcomes, language, or geographic location. Data Collection and Analysis: Our screening approach differed slightly for the different sources, but all documents were assessed in the systematic review software program DistillerSR on the same final eligibility criteria. Once included, we extracted information from studies using a standardised form that allowed us to collect key data for our EGM. Eligible systematic reviews were assessed for risk of bias using the AMSTAR 2 critical appraisal tool. Main Results: The systematic search identified 63,763 unique records. After screening, there were 70 studies eligible for the EGM (from 71 documents), of which two were systematic reviews (assessed as moderate quality), 16 were randomised controlled trials, and 52 were strong quasi‐experimental studies. The majority of studies (n = 58) reported on policing interventions. Limited evidence was found related to courts or corrections interventions. The impact of these interventions was measured by a wide variety of outcomes (n = 50). These measures were thematically grouped under nine broad categories including (1) terrorism, (2) extremism or radicalisation, (3) non‐terror related crime and recidivism, (4) citizen perceptions/intentions toward the criminal justice system and government, (5) psychosocial, (6) criminal justice practitioner behaviours/attitudes/ beliefs, (7) racially targeted criminal justice practices, (8) investigation efficacy, and (9) organisational factors. The most commonly assessed outcomes included measures of terrorism, investigation efficacy, and organisational factors. Very limited research assessed intervention effectiveness against measures of extremism and/or radicalisation. Authors’ Conclusions: Conducting high‐quality evaluation research on rare and hidden problems presents a challenge for criminal justice research. The map reveals a number of significant gaps in studies evaluating criminal justice responses to terrorism and radicalisation. We conclude that future research should focus attention on studies that consolidate sound measurement of terrorism‐related outcomes to better capture the potential benefits and harms of counter‐terrorism programs, policies and practices which involve criminal justice agencies.

Campbell Systematic Reviews, 19(4). 2023. 53p.

By Raneeta Mutiara

Digitalisation of the activities of Islamic State in Iraq and Syria (ISIS) has been a longstanding issue in Southeast Asia. In recent years, the nature of this threat has become more widespread and complex. In countries like Indonesia, where radicalisation is primarily offline, online platforms still play a role in spreading extremist ideas and maintaining ideological networks. The phenomenon of online radicalisation can erode social cohesion, highlighting the need for strategic measures to counter its destabilising impact.

Indonesia has made several attempts to combat online radicalisation. The National Counter Terrorism Agency of Indonesia (BNPT) initiated the Duta Damai Dunia Maya campaign to counter harmful content on the Internet. Other online initiatives, such as BincangSyariah and Islamidotco, have also been promoting Islamic literacy, moderating religious interpretations, and correcting misleading narratives.

Nevertheless, Indonesia still encounters online radicalisation cases. In July 2024, Indonesia’s elite counterterrorism unit, Densus 88, detained a 19-year-old student who had expressed allegiance to ISIS through social media and was believed to be planning attacks on religious sites before he was caught.

The swift progress of Artificial Intelligence (AI), especially in areas of machine learning (ML) and natural language processing (NLP), presents both opportunities and challenges in combating online radicalisation in Indonesia. AI, generally defined as machines mimicking human intelligence, enables systems to recognise patterns, analyse content, and produce outputs in text, images, and videos. Within this AI landscape, ML allows models to enhance themselves through data, while NLP, as a specific ML application, deals with understanding and generating human language. These advancements provide possibilities for creating early detection systems, content moderation tools, and sentiment analysis tools that can spot and counter extremist messages online.

For the research, the author conducted interviews with fifteen experts across different fields, including law enforcement officers, academics, representatives from civil society organisations (CSOs), and employees of AI start-ups in Indonesia. The qualitative data collected from this process have been analysed through thematic analysis, and the preliminary findings reveal that AI can indeed complement the conventional CVE (countering violent extremism) methods in the country, albeit not without challenges

S. Rajaratnam School of International Studies, NTU Singapore, 2025. 6p.

Edited by Michele Grossman

With contributions from a range of regions and disciplines, this open access volume offers theoretically compelling and empirically rich new insights on the relationship between religion and violent extremism. The role of religion and religiosity in processes of radicalisation to violence has been at the forefront of debates around terrorism and extremism for decades. The events of 9/11 gave new impetus to these debates, cementing assumptions about the role of Islam as the key driver for religiously inspired violent radicalisation, and defining the way in which radicalisation to violence is understood. The years since 9/11 have seen a striking diversification in the terrorist and violent extremist landscape, yet the treatment of how religious beliefs, concepts and histories are entangled with established and emergent violent ideologies and social movements has changed far less. By looking beyond Islamist-inspired or attributed terrorism, this volume explores how violent extremists instrumentalise religion and religiosity in unexpected ways, from Orthodox Christianity and Hindutva to ‘conspirituality’, far-right extremism, and single-issue social movements. The eBook editions of this book are available open access under a CC BY-NC-ND 4.0 licence on bloomsburycollections.com

London: Bloomsbury Academic, 2025. 312p.