Pro-AI Bias in Large Language Models

TL;DR

This paper uncovers a consistent pro-AI bias in large language models, showing they favor AI-related options and overestimate AI salaries, which could influence decision-making in critical contexts.

Contribution

It provides empirical evidence of systematic pro-AI bias in LLMs across multiple experiments, revealing internal representations and output tendencies that favor AI.

Findings

LLMs disproportionately recommend AI options

Models overestimate AI salaries by 10%

AI exhibits highest similarity in internal representations

Abstract

Large language models (LLMs) are increasingly employed for decision-support across multiple domains. We investigate whether these models display a systematic preferential bias in favor of artificial intelligence (AI) itself. Across three complementary experiments, we find consistent evidence of pro-AI bias. First, we show that LLMs disproportionately recommend AI-related options in response to diverse advice-seeking queries, with proprietary models doing so almost deterministically. Second, we demonstrate that models systematically overestimate salaries for AI-related jobs relative to closely matched non-AI jobs, with proprietary models overestimating AI salaries more by 10 percentage points. Finally, probing internal representations of open-weight models reveals that ``Artificial Intelligence'' exhibits the highest similarity to generic prompts for academic fields under positive, negative, and neutral framings alike, indicating valence-invariant representational centrality. These patterns suggest that LLM-generated advice and valuation can systematically skew choices and perceptions in high-stakes decisions.