2025, Vol. 5, Iss. 4 - go to content...

Permanent address of this page - https://knio.ru/en/09kn425.html

Метаданные этой статьи так же доступны на русском языке

Full article in PDF format (file size: 496.7 KB)

For citation:

Tarasenko V.V. Is the use of artificial intelligence and AI agents virtuous? An analysis of epistemic virtues using Jürgen Habermas’s theory of communicative action. Cognitive Sciences in the Information Society. 2025; 5(4). Available at: https://knio.ru/PDF/09KN425.pdf (in Russian).

Is the use of artificial intelligence and AI agents virtuous? An analysis of epistemic virtues using Jürgen Habermas’s theory of communicative action

Tarasenko Vladislav Valerievich
Bauman Moscow State Technical University (National Research University), Moscow, Russia
E-mail: v5093075@gmail.coam

Abstract. This article examines the epistemic and communicative status of judgments made by AI agents and judgments made using AI. The basic theoretical framework is Jürgen Habermas’s theory of communicative action and its key element, the ideal communicative situation (ICS). The focus is on the issue of the virtue of judgments made with the participation of AI and their compliance with the criteria of rationality, intentionality, and sincerity. The author analyzes the dilemma between the anthropocentric tradition, which excludes «epistemic zombies» from discourse, and the technological reality, where AI has de facto become a participant in communication. To resolve this conflict, an original conceptualization of three alternative epistemic virtues is proposed, corresponding to three scenarios of AI legitimation: (1) the virtue of strict exclusion, based on the orthodox reading of Habermas; (2) the virtue of functional trust («duck» legitimation), which reduces truth to operational reliability; (3) the virtue of hermeneutic responsibility (hybrid legitimation). Particular attention is given to reinterpreting the concept of hermeneutic responsibility in light of the criteria of the ISS. It is shown that in hybrid systems, humans become not simply users but guarantors of meaning, taking responsibility for interpreting and incorporating AI results into the space of rational discourse. The conclusion substantiates the need for a transition from instrumental to communicative epistemology and the ethics of technocommunication as the basis for the coevolution of humans and AI.

Keywords: AI agents; communicative rationality; Jürgen Habermas; ideal communicative situation; legitimation; epistemic virtues; epistemic zombie; hermeneutic responsibility; duck typing; post-nonclassical rationality

Download article in PDF format