Rothut, S., Schulze, H., Hohner, J., Greipl, S., & Rieger, D. (2026). Radicalization and the internet: 25 years of (online) radicalization research. The Communication Review, 1–45. https://doi.org/10.1080/10714421.2026.2658935
Summary
This scoping review synthesizes a quarter-century (2000–2025) of scholarship on the internet’s role in radicalization, drawing on 382 publications to map the field’s conceptual, methodological, and empirical contours. The authors advance an integrative definition of radicalization as an incremental, multi-factor process leading to the adoption of radical/extremist ideologies and/or acceptance of ideologically motivated (violent) action. Their central argument is that the internet operates as a catalyst rather than a root cause, and that strictly separating “online” and “offline” radicalization is no longer conceptually tenable in a hybrid, socio-technical digital society. To organize a fragmented interdisciplinary literature, they propose a dual analytical framework pairing macro/meso/micro levels with Lasswell’s communication-process elements (communicator, content, medium, recipient).
Key Contributions
- Broadest temporal (2000–2025) and cross-phenomenological synthesis of online radicalization research to date, covering 382 publications.
- An integrative, processual, multi-factor, multi-outcome definition of radicalization.
- A dual analytical framework combining macro/meso/micro levels with the Lasswell formula to disaggregate “the internet” into specific, analyzable factors.
- A mapping of the field’s trends, geographic and phenomenological biases, methodological trajectory, and platform coverage.
- A typology of recurring factor clusters (algorithmic curation, content strategies, communicator tactics, exposure/engagement).
- Identification of persistent gaps — longitudinal design, causal inference, understudied platforms, non-Western contexts.
Methods
A systematic scoping review spanning January 2000 to December 2025, searching Web of Science, Communication and Mass Media Complete, ACM Digital Library, the VOX-Pol library, and Google Scholar with Boolean English/German keyword combinations. From an initial 3,532 records, two-stage screening (title/abstract, then full-text) yielded 382 included publications after excluding duplicates, inaccessible, purely methodological/technical, and prevention-focused items. The corpus was quantitatively coded for field descriptives, conceptual categories, and the two analytical lenses, with qualitative extraction of findings and implications. Intracoder reliability on a 10% subsample (n=38) reached Krippendorff’s alpha of 0.90–1.00.
Findings
- 76.7% of publications were peer-reviewed journal articles, with a substantial gray-literature minority reflecting the field’s applied orientation.
- Geographic focus skews Western (UK, Germany, US most studied); over half of publications addressed radicalization in non-country-specific terms.
- Output grew sharply from 2014, peaking at 47 publications in 2025; 193 of 382 appeared in 2021–2025.
- Islamist extremism dominates (34%), followed by far-right (20.2%), with growing post-2022 attention to misogynist/Incel, QAnon/conspiracy, and anti-system radicalization.
- Empirical studies (71.7%) dominate post-2018; qualitative work leads while computational and mixed methods rise, but longitudinal studies remain rare (n=34).
- Less than half (47.4%) explicitly define radicalization, and only 15.7% offer a dedicated online-radicalization definition.
- Macro-level work (48.2%) stresses information/contact functions, algorithmic curation, and mainstreaming; meso-level (18.6%) highlights echo chambers, co-radicalization, and emotional convergence; micro-level (30.9%) points to hybrid pathways, salience for socially isolated individuals, and possibly shortened timelines.
- Platform research concentrates on Twitter/X (n=36), YouTube (n=22), Facebook (n=21), and increasingly Telegram (n=17); alternative platforms and gaming environments remain understudied.
- Recipient-level evidence indicates single short-term exposure rarely radicalizes; effects depend on prior attitudes, identification, dispositional traits, and protective factors like media literacy and democratic trust.
Connections
As a scoping review authored by an overlapping team, this note is a natural conceptual anchor for other work in the register, including Rothut2026-or. Its hybrid, socio-technical framing and attention to phenomenological differentiation (misogynist/Manosphere, conspiracy, anti-system) connects to work on specific extremist ecosystems and platform dynamics such as Askanius2026-de, Grusauskaite2026-po, and Marwick2025-ov; its emphasis on algorithmic curation and platform-specific research speaks to studies like Rieder2026-pp and Kulichkina2026-zk.
Podcast
A research-radio episode discusses this paper: 🎧 MP3 · Spotify · Apple Podcasts