AI Has Conquered Saudi Arabia’s Media Industry, But Not Equally

Artificial intelligence has swept through the world’s film sets, editing suites, and podcast studios with remarkable speed, but a new study from Saudi Arabia reveals a striking twist in that story: in some markets, the question is no longer whether media professionals will adopt AI, but how deeply, how safely, and how fairly they will do so. A cross-sectional survey of 400 audio-visual media professionals in Saudi Arabia, published in Discover Artificial Intelligence, found that 95.8 percent of respondents—383 of 400—already use AI tools in their content production work. Only 17 reported no use at all. That near-saturation changes the entire analytical picture, and the researchers behind the study argue it demands a new way of thinking about technology adoption altogether.

The study, conducted in 2025 using purposive sampling, was led by Saad Faraj Alenzi and Ahmad Tawalbeh of Gulf University in Bahrain, together with Ferhat Yilmaz and colleagues at St. Petersburg State University. Rather than treating AI adoption as a single event—a switch flipped at one moment—the team proposed what they call a Barrier–Enabler Ecosystem Model. The framework integrates three established theories: the Technology Acceptance Model, which explains how perceived usefulness and ease of use shape individual decisions; Digital Divide Theory, which describes how access to motivation, material resources, skills, and usage opportunities is unevenly distributed; and Institutional Theory, which captures the coercive, mimetic, and normative pressures that organizations and their members feel from regulation, peer behavior, and professional norms. The model treats barriers and enablers not as two separate lists but as opposing valences of the same underlying dimensions—cost is a barrier when tools are expensive and a facilitator when they are affordable, for example.

Saudi Arabia makes an unusually revealing setting for this kind of analysis. Under Vision 2030, the Kingdom has pursued an explicit, state-directed program of digital transformation and national AI capability building, complete with dedicated institutions such as the Saudi Data and AI Authority. Media organizations therefore operate under unusually visible technological expectation from the top. At the same time, the sector faces constraints that are anything but generic: compliance obligations under the Personal Data Protection Law, well-documented performance gaps in Arabic-language processing relative to English, subscription costs benchmarked against international rather than local purchasing power, and uneven distribution of training capacity across regions and employment types. Strong institutional encouragement and substantive structural constraint operate simultaneously—precisely the configuration a multi-level framework is needed to explain.

Among the small group of non-users, privacy and security concerns topped the list of barriers, reported by 12 of the 17 respondents, followed closely by high subscription costs, reported by 11. The researchers are careful to note that with only 17 responses, this ranking is descriptive and exploratory rather than generalizable to the national sector. Still, the prominence of privacy is theoretically telling. Unlike cost or skill deficits, privacy is not resolved by acquiring resources; it persists even for well-resourced practitioners. The authors draw on the privacy calculus tradition to explain why: audio-visual professionals handle disproportionately confidential material—unpublished footage, pre-release assets, client information, and in journalism, source identities. Cloud-based AI tools typically require that material to leave organizational control for processing, and the Personal Data Protection Law makes the consequences of misjudgment concrete rather than hypothetical. A practitioner may judge a tool both useful and easy to use and still decline it because the conditions of its use are deemed impermissible.

Across the full sample of 400, the most frequently endorsed facilitating condition was training programs and specialized workshops, followed by easy-to-use tools at affordable prices, and then high-quality training data and open-source resources. Technical support services, collaboration with technology companies, clear standards and ethics frameworks, and industry case studies completed the ranking. The researchers emphasize that these conditions span three distinct levels—individual capability, organizational and market provision, and sector-level normative infrastructure—suggesting that no single category of intervention will be sufficient on its own. In a sample where nearly everyone already uses AI, the appetite for training likely reflects an awareness that current use is shallower than it could be, not an absence of access.

The demographic findings add another layer of nuance. AI readiness varied significantly by age, occupational sector, geographic region, and monthly income, but not by gender or education. The occupational association was the largest in the study: government-sector respondents reported the highest readiness, while retired professionals reported the lowest, though that latter group comprised only ten people. The authors offer a theoretically motivated institutional reading: state and public media organizations operate under explicit Vision 2030 digitalization expectations, and where organizations respond by providing tools, training, and time, practitioners experience institutional pressure as tangible support. They stress, however, that institutional pressure and organizational policy were not directly measured, and alternative explanations such as selection into public-sector employment cannot be excluded.

Perhaps the most intriguing result is regional. The Central region, home to Riyadh and the bulk of the Kingdom’s media infrastructure, reported the lowest mean readiness, while the Southern region reported the highest. Rather than invoking vague cultural explanations, the researchers propose a socio-technical reading: practitioners in the densest media market have the most cumulative exposure to AI tools in real production conditions, and therefore the most direct experience of where those tools fail—particularly in Arabic contextual processing, which ranked as the second-highest technical challenge in the survey. Readiness measured as attitude may be depressed by informed appraisal rather than inexperience. Conversely, elevated readiness in the South may reflect expansion-driven demand, as Vision 2030 tourism and gigaproject development generate urgent need for promotional content that automated tooling can help meet under capacity pressure. Both readings remain hypotheses, the authors caution, since neither market maturity nor demand pressure was measured.

The Arabic-language finding deserves particular emphasis because it is specific to the linguistic context and cannot be reduced to the generic constraints documented in Anglophone research. Difficulty understanding contextual meaning in Arabic content ranked second among technical challenges, behind only the need for large volumes of high-quality training data. Arabic presents specific processing difficulties—morphological richness, diglossia between Modern Standard Arabic and regional dialects, and orthographic underspecification of vowels—that general-purpose models trained predominantly on English corpora handle unevenly. The implication is significant for policy: no amount of practitioner upskilling can compensate for a model that misreads dialectal register or fails to disambiguate meaning. The researchers recommend that national AI bodies fund the development of Arabic-language models and resources tuned to media production, including dialectal variation, transcription and subtitling accuracy, and contextual disambiguation, alongside high-quality Arabic audio-visual training corpora with clear rights provenance.

On the question of job displacement, the study offers a striking contrast with Western findings. Concern about job impact was the least frequently reported barrier among non-users, cited by just two respondents, whereas displacement anxiety is prominent among Western journalists in prior research. The authors note this divergence rests on only two responses and should be treated as indicative, but suggest that sectoral expansion under Vision 2030 may reduce the perceived plausibility of displacement, or that differences in labor market structure may alter its perceived consequences. Respondents’ expectations about AI’s future impact were oriented toward efficiency first and creative augmentation second, indicating a complementarity rather than substitution view of the technology.

The study’s limitations are stated with unusual candor. The cross-sectional design precludes causal inference; the purposive sample is not statistically representative of the national workforce; the barrier profile rests on just 17 responses; and the attitudinal readiness composite was assessed for internal consistency only, without confirmatory factor analysis or test–retest validation. The proposed ecosystem model was not statistically estimated and awaits empirical validation through methods such as structural equation modeling with measured institutional and organizational variables. Within those bounds, the conclusions are clear: in near-saturated markets, advancing AI integration depends less on persuading professionals to adopt—most already have—than on resolving the data governance, cost, capability, and Arabic-language conditions that determine how well they are able to do so. The productive question, the authors argue, is no longer incidence of adoption but the depth, quality, and equity of integration.

Subject of Research: AI adoption readiness, barriers, and facilitating conditions among audio-visual media professionals in Saudi Arabia

Article Title: Artificial intelligence readiness among audiovisual media professionals in Saudi Arabia and its barriers and facilitating conditions

Article References: Alenzi, S. F., Tawalbeh, A., Yilmaz, F., Akhmatshina, E. K., & Nikonov, S. B. (2026). Artificial intelligence readiness among audiovisual media professionals in Saudi Arabia and its barriers and facilitating conditions. Discover Artificial Intelligence, 6(1), Article 1378. https://doi.org/10.1007/s44163-026-02231-x

Image Credits: AI Generated

DOI: 10.1007/s44163-026-02231-x

Keywords: artificial intelligence, Saudi Arabia, audio-visual media, technology acceptance, digital divide, institutional theory, Vision 2030, privacy, Arabic-language AI, media production, training, adoption barriers