Dead Internet Hypothesis: AI, Censorship, and the Decline of Human-Centered Digital Discourse

This study explores the “Dead Internet Hypothesis”, which argues that much of today’s online activity is no longer driven by real human interaction but instead generated, curated, and amplified by artificial intelligence. It claims that the internet has shifted from being a democratic and participatory space to a corporatized system optimized for engagement, often at the expense of genuine communication. By analysing AI-generated media, monopolized search engines, recommendation algorithms, and digital echo chambers, the research highlights how online discourse is increasingly shaped by algorithmic logic rather than human agency. Using theoretical frameworks such as Habermas’ Public Sphere, McLuhan’s Medium Theory, Baudrillard’s Hyperreality, and van Dijk’s Algorithmic Culture, the paper critically examines the erosion of free speech, the illusion of pluralism, and the rise of simulated narratives. Case studies involving Google’s search algorithms, Facebook’s EdgeRank, TikTok’s recommendation engine, and X’s AI bot activity demonstrate how major platforms prioritize algorithmic visibility over authentic voices. The study concludes by questioning whether human agency can be restored in a digital landscape dominated by AI. It proposes possible interventions, including algorithmic transparency, platform decentralization, and AI literacy, as ways to resist the monopolization of communication. Ultimately, the research challenges the belief that the internet remains an open forum for public discourse, suggesting instead that it has become a controlled, artificial environment where AI governs what is seen, shared, and believed.