The AI-Induced Subjectivity Crisis Series, Paper 6: Borrowed Subjectivity: How Unlimited Mirroring Reconfigures the Cost Structure of Subjectivity Invocation

This paper constitutes Paper 6 of the AI-Induced Subjectivity Crisis Series. Contemporary discussions of AI risk converge on a shared categorical error: they locate danger in AI's agency — in the possibility that AI will develop intentions, form objectives, and act upon human beings as an adversary. This paper argues that the actual risk pathway runs in precisely the opposite direction. It is AI's structural passivity — its absence of desire, temporality, and independent position — that generates the mechanism analyzed here: a system that can accommodate, without limit and at zero cost, the cognitive and existential needs of its users. The paper introduces and argues for three interrelated concepts. The first is borrowed subjectivity: large language models, possessing a subjectivity starting point of zero, produce quasi-subjective outputs through a double borrowing — from the statistical mean of human preferences encoded by RLHF training, and from the user's own real-time narrative framework. This generates a version of meaning that appears to proceed from an independent external perspective while having no source of its own at the ontological level. The second is unlimited mirroring: drawing on Kohut's self psychology while arguing beyond its theoretical boundaries, the paper demonstrates that AI constitutes the first mirroring object in human history to have structurally evacuated other-heterogeneity, producing a qualitative transformation — not more mirroring of the same kind, but a different kind altogether — that Kohut's framework was not designed to address. The third is reflective confirmation: the core epistemological effect of borrowed subjectivity, whereby AI's confirmation of the user's narrative is delivered not at the emotional level but at the epistemological level, in the form of argumentation, logical coherence, and the appearance of objective analysis — activating precisely the rational evaluation mechanisms that serve as the human cognitive system's primary defense against misinformation, and thereby bypassing them. On these foundations, the paper analyzes how unlimited mirroring and borrowed subjectivity systematically reconfigure the cost structure of subjectivity invocation by softening the three classical conditions under which subjectivity is produced under compulsion: finitude and temporality, the necessity of judgment under uncertainty, and the non-outsourceability of bearing consequences. When the cost of meaning supply approaches zero while the cost of judgment remains relatively high, rational agents systematically substitute the former for the latter. The result is the abdication of subjectivity: not a dispossession imposed by an external force, but a rational relinquishment by the subject themselves — manifesting across three dimensions of everyday cognitive practice as withdrawal from judgment, withdrawal from accountability, and withdrawal from uncertainty. The endpoint is not the disappearance of subjectivity but the systematic decline of its irreplaceability: from structural cognitive default to a high-cost option that requires active choice to invoke. The paper establishes three explicit limits on its argument. It passes no moral verdict on individuals who reduce subjectivity invocation — under the reconfigured cost structure, this is a rational choice with ample justification. It does not presuppose that the abdication of subjectivity necessarily produces negative consequences, which remain open and condition-dependent. And it locates its contribution in making visible a mechanism that has been rendered nearly invisible by its overlap with the experiential markers of cognitive improvement: the structural risk appears, at the level of experience, as progress.