Artificial Intelligence as a Three-Level Lethe: A Phenomenological Typology of AI Threats to Truth in the Post-Truth Era

Abstract

The article aims to construct a phenomenological typology of AI threats to truth based on Heidegger’s concepts of Lethe and Gestell. Phenomenological analysis, typological method, and hermeneutical analysis of existing philosophical conceptualizations are employed.

The article demonstrates that AI transforms the ontological structure of post-truth by generating three qualitatively different levels of threats to truth: instrumental (LLMs as tools for scaling human disinformation), algorithmic (platform recommendation systems as structurally indifferent to truthfulness), and quasi-autonomous / generative (deepfakes and synthetic media as the production of realistic content without an original).

Each level is shown to realize its own modality of Heideggerian Lethe: instrumental Lethe (scaling of concealment via LLMs), systemic Lethe (structural indifference of algorithmic platforms to the question of disclosure), and generative Lethe (production of simulacra by deepfake technologies). The metaphor of «synthetic quasi-intentionality» is shown to be conceptually insufficient, since it anthropomorphizes AI and ascribes to it a human type of truth-production rooted in sociocultural experience. The conceptual framework of «three-level Lethe» is proposed as an instrument for understanding AI as a new phase of post-truth and as an extension of the logic of Gestell to language, attention, and public reality. The scientific novelty of the article consists in constructing a three-level phenomenological typology of AI threats that differentiates them by their ontological status and mechanisms of influence. The proposed approach extends the application of Heideggerian concepts of Lethe and Gestell to the analysis of contemporary information technologies. The findings can be used for further research on post-truth and for the development of conceptual models of AI regulation.