Is AI-Generated Extremism Credible? Experimental Evidence from an Expert Survey

Rapid progress in Artificial Intelligence (AI) over the past few years has unfolded important questions—and triggered anxiety—about the potential (mis)use of AI-generated content in information warfare and disinformation campaigns. A crucial question remains unanswered, however: whether—and to what...

Descripción completa

Guardado en:  
Detalles Bibliográficos
Autores principales: Baele, Stéphane J. (Autor) ; Naserian, Elahe (Autor) ; Katz, Gabriel (Autor)
Tipo de documento: Electrónico Artículo
Lenguaje:Inglés
Publicado: 2025
En: Terrorism and political violence
Año: 2025, Volumen: 37, Número: 8, Páginas: 1060-1076
Acceso en línea: Volltext (lizenzpflichtig)
Verificar disponibilidad: HBZ Gateway
Palabras clave:

MARC

LEADER 00000caa a22000002c 4500
001 1941667880
003 DE-627
005 20251211143636.0
007 cr uuu---uuuuu
008 251120s2025 xx |||||o 00| ||eng c
024 7 |a 10.1080/09546553.2024.2380089  |2 doi 
035 |a (DE-627)1941667880 
035 |a (DE-599)KXP1941667880 
040 |a DE-627  |b ger  |c DE-627  |e rda 
041 |a eng 
084 |a 2,1  |2 ssgn 
100 1 |a Baele, Stéphane J.  |e VerfasserIn  |0 (DE-588)1150539364  |0 (DE-627)1010814842  |0 (DE-576)497068648  |4 aut 
109 |a Baele, Stéphane J. 
245 1 0 |a Is AI-Generated Extremism Credible? Experimental Evidence from an Expert Survey 
264 1 |c 2025 
336 |a Text  |b txt  |2 rdacontent 
337 |a Computermedien  |b c  |2 rdamedia 
338 |a Online-Ressource  |b cr  |2 rdacarrier 
520 |a Rapid progress in Artificial Intelligence (AI) over the past few years has unfolded important questions—and triggered anxiety—about the potential (mis)use of AI-generated content in information warfare and disinformation campaigns. A crucial question remains unanswered, however: whether—and to what extent—this technology can be harnessed in a credible way by extremists, whose language is usually characterized by highly specific subcultural registers that make them unlikely to be easily mimicked by AI models. This paper answers that question, offering the first rigorous evaluation of the credibility of synthetic (AI-generated) extremist prose. Using an expert survey experiment measuring the credibility of fake incel forum posts and ISIS magazines paragraphs generated through an original workflow for the production of synthetic extremist text, we show that these texts have high credibility scores, confusing world-leading experts. These findings, which are discussed under the broader light of the emerging literature on nefarious “dual-use” of synthetic content, not only define and evaluate the threat of extremists harnessing language models for propaganda, but they also more broadly strengthen our understanding of the role language models are due to play in hostile political communication, and ultimately resurface the debate on the human costs of technological progress. 
650 4 |a Islamic State 
650 4 |a incel 
650 4 |a Language 
650 4 |a Experiment 
650 4 |a language models 
650 4 |a Extremism 
650 4 |a Artificial Intelligence 
700 1 |a Naserian, Elahe  |e VerfasserIn  |0 (DE-588)1384091084  |0 (DE-627)1945913940  |4 aut 
700 1 |a Katz, Gabriel  |e VerfasserIn  |0 (DE-588)1047975637  |0 (DE-627)779661966  |0 (DE-576)401850005  |4 aut 
773 0 8 |i Enthalten in  |t Terrorism and political violence  |d London : Cass, 1989  |g 37(2025), 8, Seite 1060-1076  |h Online-Ressource  |w (DE-627)349234795  |w (DE-600)2080322-9  |w (DE-576)113563507  |x 1556-1836  |7 nnas 
773 1 8 |g volume:37  |g year:2025  |g number:8  |g pages:1060-1076 
856 4 0 |u https://doi.org/10.1080/09546553.2024.2380089  |x Resolving-System  |z lizenzpflichtig  |3 Volltext  |7 1 
912 |a NOMM 
935 |a mkri 
951 |a AR 
ELC |a 1 
LOK |0 000 xxxxxcx a22 zn 4500 
LOK |0 001 4809080927 
LOK |0 003 DE-627 
LOK |0 004 1941667880 
LOK |0 005 20251211143636 
LOK |0 008 251120||||||||||||||||ger||||||| 
LOK |0 035   |a (DE-2619)KrimDok#2025-11-19#5C893B3D3FD135AA7D762311495DDC399C35E855 
LOK |0 040   |a DE-2619  |c DE-627  |d DE-2619 
LOK |0 092   |o n 
LOK |0 852   |a DE-2619 
LOK |0 852 1  |9 00 
LOK |0 935   |a zota 
LOK |0 939   |a 20-11-25  |b l01 
ORI |a WA-MARC-krimdoka001.raw