Bots, Trust, and Generative AI: Framing Communicative Agency in a Small-Language Media Context
Abstract
This article examines how bots, chatbots, and generative AI are represented as forms of communicative agency in a small-language media environment. It draws on a mixed content analysis of 195 digital news items published by 1TV.ge, Euronews Georgia, Imedi News, and Rustavi 2 between September 15, 2024, and December 31, 2025, a fifteen-and-a-half-month period spanning parts of two calendar years. Rather than evaluating AI systems technically or measuring audience responses, the article analyzes how news coverage assigns roles to automated and AI-based systems through publication patterns, outlet profiles, dominant frames, functional emphases, trust narratives, risk discourse, media literacy elements, and named AI systems. The corpus was most often organized through a technological or business news frame (93 items; 47.7%). Other frames included myth-nightmare (36; 18.5%), media literacy/fact-checking (26; 13.3%), global political competition/information war (24; 12.3%), and myth-idyll (16; 8.2%). The comparison shows that Euronews Georgia emphasized international and security-related coverage, Imedi News foregrounded education and textual interaction, Rustavi 2 emphasized consumer-facing products and visual applications, and 1TV.ge displayed a comparatively balanced explanatory profile. The article argues that bot transformation is better understood as a change in communicative roles than as a simple replacement of older automation by AI chatbots. In this process, agency, usefulness, trust, risk, and responsibility intersect. The Georgian-language case contributes to Human-Machine Communication, AI-mediated communication, and media-framing research by showing how global AI debates are localized outside dominant English-language media environments. Coding was conducted by a single researcher; an intracoder stability check on 20 items (approximately 10% of the corpus) produced no changes to dominant-frame assignments.
