LLMs as Oracles: Reliance on LLMs for Subjective Personal Questions

arXiv:2609.14849v1 Announce Type: new Abstract: We characterize how people are turning to LLMs as oracles: all-knowing authorities on subjective personal questions. Motivated by risks to users' autonomy and well-being, we develop a typology and LLM-based methods to measure this form of AI reliance at scale and understand how people are offloading judgment and decision-making to AI. Applying our typology to public usage data (68K prompts from WildChat and ThoughtTrace), we find that LLM-as-oracle use has increased over time (2023-2026) and is more prevalent among younger users. We further build a privacy-preserving data donation tool to analyze individuals' longitudinal usage data (140K prompts from 52 participants), identifying similar trends. People are often unaware of their own LLM-as-oracle use, and express dissatisfaction with this behavior after seeing our tool's analysis. Finally, we identify two drivers of LLM-as-oracle use: people's perceptions of AI and the behavior of AI models themselves, which motivate possible interventions to support users' self-deliberation.

Risks to autonomy and well-being from LLM reliance, but potential for interventions.

The study finds increasing reliance on LLMs as oracles for subjective personal decisions, potentially undermining user autonomy and well-being. The lack of user awareness and expressed dissatisfaction suggests negative effects on personal judgment. However, the situation is not extreme or universal yet, and potential interventions are noted.

This is empirical research documenting a real and growing behavioral pattern of offloading personal judgment to LLMs, with evidence of user unawareness and dissatisfaction once informed. The implication is materially concerning for autonomy, though the paper also proposes interventions, and sample sizes for the longitudinal analysis are modest.

The study provides evidence of increasing reliance on LLMs for subjective personal questions, leading users to offload judgment and decision-making. This trend, especially among younger users, poses a significant risk to individual autonomy and well-being. The finding that users are often unaware and later dissatisfied highlights a concerning, subtle erosion of human agency.

The study shows increasing reliance on LLMs for subjective personal questions, which can erode autonomy and well-being. The trend is stronger among younger users, who may be more vulnerable. Users' lack of awareness and subsequent dissatisfaction indicate a negative impact.

SOURCE     arXiv cs.CY (Computers and Society) (tier 1)
   ↓
DOCUMENT   c0712757-212c-47d9-b059-2bce93fb2d21
           https://arxiv.org/abs/2609.14849
   ↓
EVIDENCE   3 extracted excerpts
   ↓
MODEL RUN  5 runs, methodology 1.0.1
   ↓
SCORE      -56  (Adverse)
   ↓
CONFIDENCE 80%