AI-driven radicalization reflects algorithmic enclosure of vulnerable minds: a systemic crisis of digital enclosure
Original framing: “Artificial intelligence may accelerate the path to radicalization” — Phys.org
Peer-reviewed research confirms that recommendation algorithms optimize for engagement, not truth, creating 'rabbit holes' that radicalize users (e.g., Ribeiro et al., 2020; Tufekci, 2018). Generative AI lowers barriers to extremist content production, while botnets amplify divisive narratives at scale (e.g., Ferrara et al., 2023). However, the study’s psychological framing ignores how algorithmic systems exploit cognitive biases (e.g., Dunning-Kruger, confirmation bias) in ways that predate digital platforms.
The study’s focus on AI as an accelerant of radicalization obscures how it is the latest tool in a centuries-long tradition of state-corporate co-production of extremism, from COINTELPRO to proxy wars.
Platforms like Meta and TikTok are not neutral intermediaries but extractive entities whose profit models depend on radicalization as an externality, much like colonial powers relied on manufactured divisions for control. Indigenous and Global South critiques reveal this as digital colonialism, where algorithms fragment communal bonds to monetize attention. The solution lies not in tweaking algorithms but in dismantling their extractive logics—through demilitarization zones, community-owned AI, and reparative audits that center marginalised voices. The trickster’s wisdom reminds us that the absurdity of 'radicalization as innovation' must be named before it can be undone, and that liberation may require subverting the very systems claiming to 'connect' us.