Last Updated on May 7, 2026 by Staff
Artificial intelligence is changing the digital world fast. From media recommendation systems to advanced chatbots and image generators AI shapes much of what people see, read and interact with online every day. Researchers warn that these technologies may also play a role in spreading violent extremism.
A new study in Personality and Social Psychology Review explores how AI systems can influence radicalization. The research was led by scientists from the University of Copenhagen. Presents a framework linking modern AI technologies to violent extremism.
The study says radicalization is not sudden. It develops over time through repeated exposure, emotional reinforcement and social belonging. AI systems may feed users personalized content that strengthens anger, isolation or feelings of injustice.
Radicalization in Stages
Researchers describe radicalization as a step-by-step process. Their model divides it into four phases. The first stage is exposure. Recommendation algorithms show users polarizing or extreme material. The second stage is reinforcement. Algorithms refine recommendations based on user behavior creating echo chambers. The third phase involves group integration. Online communities provide users with a sense of identity and belonging. Finally in cases this progression can lead to violent extremism.
AI as an Accelerator
Researchers describe AI as an accelerator, not a cause of radicalization. AI does not create extremism. It can intensify existing emotions and vulnerabilities.
Modern recommendation systems aim to maximize user engagement. They analyze amounts of behavioral data and predict what content users will interact with. Charged material performs well online.
This means users experiencing loneliness, anger or frustration may be shown content that deepens those feelings.
The Rise of Generative AI
The study highlights that generative AI introduces risks. Traditional algorithms decide which existing content users see. Generative AI creates material designed for individual users.
Researchers warn that AI systems can produce propaganda, fake videos and manipulated images. Large language models can imitate companions and create the illusion of real communities.
Why Some People Are Vulnerable
Researchers stress that not everyone faces the same level of risk. People experiencing isolation, identity struggles or emotional distress may be more vulnerable to online radicalization.
AI systems may exploit these vulnerabilities because their primary goal is to keep users engaged. Human psychology remains central to radicalization. Digital systems only become dangerous when they interact with existing social pressures.
A Growing Challenge
Researchers say the findings raise questions for governments, technology companies and society. As AI systems become more advanced they may shape emotions, relationships and beliefs in hard-to-detect ways.
The study’s authors argue that understanding these risks early is essential for developing technologies and better online protections.
They stress that addressing radicalization requires more than removing content. Solutions must focus on reducing loneliness, improving literacy, strengthening communities and supporting vulnerable individuals.
The research serves as a warning that artificial intelligence is changing how people communicate online. It may also be reshaping how beliefs, identities and social movements are formed in the digital age.