Linguistic Quality and Perceived Authenticity of AI-Generated Arabic Content: Drivers of Consumer Trust and Purchase Intention in Algeria

Abstract

As generative AI transforms Arabic marketing, understanding how linguistic quality and perceived authenticity influence consumer trust and purchase intention is critical. This study presents the first large-scale empirical investigation of the Algerian digital market, a context marked by Arabic-French bilingualism and dialectal diversity. Grounded in the Elaboration Likelihood Model (ELM), Technology Acceptance Model (TAM), and cultural congruence theory, we test a sequential mediation model where linguistic quality drives purchase intention via perceived authenticity and cognitive/affective trust.

Data from 523 Algerian consumers across six regions were analyzed using Structural Equation Modeling (SEM) and bootstrap mediation (5,000 resamples). Results show linguistic quality strongly predicts perceived authenticity (β = 0.629, p < .001), which sequentially enhances cognitive (β = 0.567) and affective trust (β = 0.535), explaining 44.2% of purchase intention variance. Integrating recent insights on AI disclosure and Arabic large language models, these findings offer vital implications for brands deploying AI-generated content in North African digital ecosystems.