While the anthropomorphic interaction enabled by artificial intelligence (AI) enhances user experience, it is also subtly reshaping how humans perceive and understand machines. Accurately recognizing AI’s instrumental nature, therefore, does not diminish the value of technology; rather, it is an intellectual prerequisite for preventing cognitive alienation in the face of the complex technologies humans themselves have created.

Instrumental nature underlying anthropomorphic interaction

Human–AI interaction is increasingly incorporating sophisticated anthropomorphic features. Functions such as natural language dialogue, human-like voices, and emotional responses can precisely trigger deeply rooted social response mechanisms in the human mind, making users more likely to form expectations of personhood toward machines displaying human-like characteristics. However, anthropomorphic forms of interaction do not alter the system’s intrinsic attributes.

A theoretical examination of anthropomorphic interaction reveals that AI’s instrumental nature remains unchanged. American philosopher John Searle’s famous “Chinese Room” thought experiment likewise suggests that even if an intelligent system can produce the appearance of “understanding” through rule-based symbol processing, it possesses neither subjective cognition nor any state of consciousness of its own. This indicates that, regardless of how AI’s modes of interaction evolve, its fundamental nature remains that of a tool. This characterization does not underestimate technological achievements; rather, it reflects a proper understanding of AI’s essential nature.

Within the international community, an increasingly clear consensus has emerged on this issue. UNESCO’s Guidelines on the Use of AI Systems in Courts and Tribunals makes clear that AI systems should serve as supportive tools rather than substitutes for legal reasoning or human judgment. At the 2026 World Digital Education Conference, UNESCO Director-General Khaled El-Enany stressed that AI must collaborate and co-create with teachers, strengthening rather than replacing their professional roles. During UNESCO’s “AI and Culture” Futures Dialogues Webinar this year, many scholars similarly argued that AI should function as a complementary tool that enhances human creativity rather than replacing the humanistic elements at the center of cultural expression. Taken together, these authoritative assessments from key fields including the judiciary, education, and culture reaffirm AI’s proper role as a technological tool rather than a bearer of responsibility.

Weakening of human subjectivity in human–machine interaction

The most serious risk posed by anthropomorphic interaction lies not in technology itself, but in the way such interaction blurs and reshapes the perceptual boundary between humans and machines. American philosopher of science and technology Don Ihde distinguishes four forms of human–technology relations.

In embodiment relations, technology is integrated into bodily experience, as in the case of eyeglasses and hearing aids, enabling humans to perceive the world through technology. In hermeneutic relations, technology provides representations of the world, such as maps and clocks, allowing humans to understand the world by interpreting technological representations. In background relations, technology is embedded within the environment, as exemplified by the internet and electricity systems, shaping the conditions in which human life unfolds. In alterity relations, technology appears as a quasi-independent entity capable of interacting with humans.

Anthropomorphic interaction in generative AI is now systematically shifting human–machine relations from Ihde’s first three forms toward “quasi-otherness” or even genuine “otherness,” making it all too easy for users to unconsciously cede their agency in making judgments. This shift arises from the combined effects of interaction design, social psychological mechanisms, and gaps in institutional norms, creating a hidden yet serious risk of diminished human subjectivity in human–machine interaction.

When tools are granted the relational status of a “quasi-other” or even an “other” through anthropomorphic interaction, what becomes blurred is not only the boundary between humans and machines, but also the very basis on which humans understand their place and affirm their own subjectivity. In this sense, accurately grasping the instrumental nature of AI is not only an epistemological issue but also a practical one. It bears directly on whether humans can, in the face of technology, uphold certain fundamental principles: remaining the source of meaning, the arbiter of value, and the bearer of responsibility.

Zhang Maocong is a professor at Shandong Normal University.