The Language of Artificial Intelligence: How Linguistic Framing in AI Systems Influences Business Decision-Making and Performance Outcomes

Abstract

This study investigates the influence of linguistic framing in AI-generated outputs on human decision-making within business contexts, employing a mixed-methods experimental design. Quantitative results revealed that framing polarity significantly affects choice consistency, risk preference, and decision accuracy, with positive framing yielding higher consistency and confidence calibration scores. Lexical certainty was positively correlated with perceived credibility and trust ratings, while explanatory depth enhanced interpretability and user confidence. Response times increased under complex framing, indicating cognitive load sensitivity. Qualitative analysis further illuminated how tone and syntactic structure shaped perceptions of AI reliability, with emotionally charged or ambiguous framing introducing interpretive bias. Hybrid visualizations confirmed that framing polarity and lexical features jointly modulate decision outcomes, underscoring the persuasive power of language in AI systems. These findings highlight the necessity for transparent prompt engineering and robust validation frameworks to mitigate hallucinations and bias in AI-assisted decision-making. The study concludes that linguistic framing is not merely a stylistic feature but a strategic variable that can significantly alter business judgments, emphasizing the need for ethical oversight and human interpretive awareness in deploying large language models for high-stakes analytical tasks.

Keywords : linguistic framing, AI decision-making, lexical certainty, interpretive bias, trust calibration, framing polarity.