Extended Abstract

Introducion: The contemporary media ecosystem is undergoing a fundamental transformation driven by the rapid advancement of Artificial Intelligence (AI). Communication theorists often describe the integration of algorithms into newsrooms as the "Third Revolution" in journalism, a paradigm shift redefining long-standing concepts such as objectivity, gatekeeping, and professional credibility. Global empirical evidence indicates that over 87% of news organizations worldwide currently utilize AI across various operational processes. In this context, the Islamic Republic of Iran Broadcasting (IRIB) News Agency, as the largest state-governed news entity in Iran, stands at a historical crossroads. The transition to intelligent automation redefines the journalist's role from a traditional "content producer" to a "content architect." However, this professional identity shift is frequently met with multi-layered psychological and structural resistance.

Viewed through the lens of Bourdieu’s Field Theory and the concept of "Habitus," any alteration in production tools challenges the established balance of power and symbolic capital within a professional field. Furthermore, integrating AI into a sovereign institution like the IRIB involves complex ethical and security variables. While international studies primarily highlight concerns regarding the "loss of creativity," resistance within sovereign media often stems from "organizational inertia," structural rigidity, and acute anxieties over data security. The primary problem addressed in this study is the profound conflict between the computational logic of algorithms and the unwritten ethical rules (Doxa) of human gatekeeping. Therefore, this research aims to conduct an in-depth exploration of the drivers of acceptance and the roots of resistance toward AI technology among IRIB News Agency staff, ultimately seeking to develop a localized model for algorithmic governance.

Methods: This study utilizes an applied, qualitative, and exploratory research design. The research population consisted of experts and employees at the IRIB News Agency. Using purposive and snowball sampling strategies, 15 participants—including senior managers, editors, journalists, and lead technical engineers—were selected for in-depth, semi-structured interviews. The data collection process continued until theoretical saturation was achieved.

The interview transcripts were systematically evaluated using an inductive thematic analysis approach. This rigorous process involved three levels of coding: 1) Basic Coding: identifying foundational semantic units; 2) Organizing Themes: categorizing initial codes based on conceptual overlaps; and 3) Global Themes: formulating overarching theoretical categories. To ensure the quality, scientific rigor, credibility, and trustworthiness of the research findings, the researchers strictly applied the four criteria established by Lincoln and Guba (1989), incorporating techniques such as member checking and external peer auditing.

Results: The thematic analysis yielded 52 basic codes, 8 organizing themes, and 4 global themes, illuminating the complex dialectic of technological acceptance and resistance within the newsroom.

Regarding the Drivers of Acceptance, the findings reveal two primary global themes: "Socio-Individual Dynamics" and "Professional Productivity." Adoption is significantly propelled by "individual innovativeness" and a generational shift, wherein younger journalists perceive AI as a cognitive extension. Furthermore, mastering AI has emerged as "symbolic capital," conferring professional prestige. Journalists actively leverage AI to overcome writer's block and automate routine tasks, thereby achieving greater independence in multimedia production.

Conversely, the Roots of Resistance are structured around "Techno-Psychological Barriers" and "Structural and Security Challenges." Resistance is exacerbated by the "black-box" opacity of AI, which generates cognitive friction. From an institutional perspective, resistance in this sovereign media environment is distinctly characterized by "environmental-normative inertia." Unlike global trends, resistance here is deeply intertwined with the fear of sensitive data leakage, fostering a restrictive climate of "information quarantine." Moreover, the "ease-of-use paradox" is starkly evident: macro-infrastructural limitations, such as internet filtering, severely diminish the perceived usefulness and adoption rates of inherently user-friendly AI tools.

The synthesis of these opposing centrifugal (drivers) and centripetal (barriers) forces culminated in the emergence of the core theme: "Protected and Pragmatic Integration." This indicates that the organization adopts a negotiated, cautious stance. Employees enthusiastically utilize AI for support, data gathering, and editorial refinement, but they strongly resist its intrusion into the "hard core" of journalistic judgment and sovereign gatekeeping.

Discussion: This research confirms that the IRIB News Agency is navigating a highly complex "cybernetic-social" transformation. Analyzed through the Social Construction of Technology (SCOT) framework, AI possesses high "interpretive flexibility" within the organization. While technical engineers view it as a neutral efficiency tool, the editorial body interprets it through a fluctuating lens, seeing it simultaneously as a "smart assistant" and a "disruptive rival." Aligning with Bourdieu’s theory, the observed resistance is not merely a skills deficit; rather, it is a protective mechanism employed by the traditional journalism habitus against reductionist algorithmic logic.

The theoretical contribution of this study lies in highlighting the fundamental distinctions in AI adoption patterns between Western commercial media and sovereign media in the Global South. While global surveys (e.g., Statista, 2025) cite the erosion of critical thinking as the primary barrier, resistance at the IRIB is profoundly tied to infrastructural macro-governance and data sovereignty concerns.

Ultimately, the study concludes that a successful transition to a "smart newsroom" requires moving beyond deterministic automation toward a paradigm of "Conjoined Agency," predicated on the "Human-in-the-Loop" (HITL) model. To achieve this, media policymakers must implement three strategic initiatives: 1) Cognitive disambiguation through hands-on, operational training; 2) The development of a secure, localized AI ecosystem to bypass the ease-of-use paradox and guarantee data sovereignty; and 3) The redefinition of job security by empowering journalists as high-level "content strategists." Only by effectively redistributing agency—delegating computational speed to algorithms while retaining ethical and sovereign judgment for humans—can the national media preserve its authoritative position in the algorithmic era.