Artificial Intelligence and Algorithmic Radicalisation: The Threat of "Synthetic Social Contagion"

Artificial Intelligence and Algorithmic Radicalisation: The Threat of "Synthetic Social Contagion"

Şenay İMAMOĞLU

Counter-terrorism and digital security strategies have extensively focused on blocking and uncovering symbols, texts of propaganda, or even videos of terrorist groups. Traditional content moderation relies upon filtering specific keywords and consuming natural language processing (NLP) models. However, the major proliferation of Generative AI (GenAI) technologies has allowed asymmetric threat actors an extraordinary capability in the spectrum of digital psychological operations (PsyOps); synthetic social contagion.

Next-generation asymmetric threat actors no longer solely spread unilateral messages of propaganda; they employ AI-based autonomous bot systems capable of creating persuasive, in-depth, and most importantly, human-like interactions with target audiences. These bots influence social learning processes, which are considered one of the most fundamental concepts in social psychology. A susceptible or even seeking individual, trusting that they are in communication with a real identity and community that truly supports, understands and guides them in the digital area, is actually being influenced within an entirety of algorithmic ecosystem. Artificial intelligence can accomplish a tailored radicalisation process in correspondence to the needs and vulnerabilities of individuals. These processes also include psychological profiling as well as personalised language patterns.

This authenticity must fundamentally modify the paradigm of cyber defence and national security construction. The issue is no longer simply barring harmful content, but rather differentiating between organic human interaction and synthetic interaction and engagement in the digital sphere. Content-centric defence mechanisms are subject to collapse in the light of this volume and speed.

Strategic Solution and Network Analysis

To defend the socio-psychological establishment from terrorism and dismantle these artificial ecosystems, the security administration must concentrate on structural network analysis, instead of content moderation. Analysts must identify the metadata mobility across the network, digital behavioural patterns and communication instead of solely focusing on what is said within the communication. The anomalies of artificial social connections created by autonomous bots must be identified using big data analytics, neutralising the danger from the core before it can exceed.

As a strategic and consistent note, the future of national security relies upon establishing algorithmic sovereignty and improving digital and social immunity against artificial social stimulations.