ИИ-аутсорсинг знаний и профессиональное становление иностранных студентов-журналистов в московских университетах

Раздел: Журналистское образование

В статье анализируется использование возможностей искусственного интеллекта в повседневной учебной практике иностранных студентов, обучающихся в Москве. На примере студентов факультетов журналистики МГУ имени М. В. Ломоносова и РУДН рассматривается, каким образом цифровое посредничество помогает им преодолевать языковые трудности, готовиться к занятиям, изучать новостную повестку, выполнять институциональные требования. Эмпирическую основу составили включенное наблюдение, а также глубинные интервью со студентами МГУ и РУДН и с руководителями факультетов. Результаты показывают, что передача ИИ задач генерации идей, перевода и написания текстов стала распространенной стратегией преодоления языковой неуверенности и нехватки времени. Обеспечивая краткосрочную эффективность, эта практика ослабляет рефлексию и чувство профессиональной ответственности. Преподаватели подчеркивают важность критического отношения к полученным результатам, необходимость проверки фактов и развития дисциплины письма. Делается вывод, что цифровая грамотность в журналистике должна пониматься – и обучающими, и обучаемыми – как этическое обязательство по развитию самостоятельного мышления.

Introduction

Artificial intelligence has transformed how university students complete academic and creative work, increasingly functioning as an assistant that organizes ideas, drafts text, and edits language. Beyond technical change, this shift reflects a broader transformation in learning practices: students may delegate aspects of intellectual effort to digital systems, turning learning into a form of knowledge outsourcing (Yan, Greiff, Teuber, Gašević, 2024). This process has implications for authorship and responsibility, particularly in fields where writing shapes professional identity. In journalism education, these dynamics are especially visible. Writing, verification, and authorship are not only academic tasks but core elements of professional legitimacy. For international journalism students in Russia, this issue is further shaped by language, as proficiency in Russian and English functions as professional capital and a pathway to employment (Murafa, Hoc, Jingyu, 2025). While AI tools can ease linguistic barriers and support writing, their use may also influence how students engage with authorship and develop confidence in their own expression.

This study examines how journalism international students in Russia experience AI-mediated knowledge outsourcing in everyday learning practices. Focusing on students’ use of AI for drafting, translation, and revision, it explores how such practices are shaped by linguistic challenges, institutional expectations, and peer dynamics. While students often describe AI as efficient and supportive, its sustained use may be associated with shifts in reflection, authorship, and perceived ownership of ideas (Shesterkina, Krasavina, 2021).

Existing research on automation in journalism education has primarily focused on curriculum reform, newsroom innovation, and professional ethics, while studies on AI in higher education often emphasize skills and responsibility frameworks. However, less attention has been paid to how AI-mediated writing practices shape students’ engagement, confidence, and sense of authorship, particularly in multilingual contexts. This study addresses this gap by examining how knowledge outsourcing relates to professional formation in journalism education.

The study addresses two key questions:

1. What experiences and concerns do journalism international students associate with AI-mediated knowledge outsourcing?

2. How can journalism education respond to this challenge through new teaching ideas and curriculum design that rebuild students’ sense of critical authorship?

In doing so, this research extends existing discussions of AI in education by foregrounding international students, non-native language environments, and the symbolic value of linguistic labor. It highlights the need to reconsider AI literacy frameworks not only as competency models but also as ethical and professional socialization processes within global journalism education. By linking AI dependency to the moral economy of journalistic labor, the study contributes to broader debates on how algorithmic mediation reconfigures professional legitimacy in transnational education.

Literature Review

Technological Change, Writing, and Professional Formation in Journalism Education

The rapid spread of automation has reshaped global discussions on journalism education, yet technology is often treated as a neutral instrument rather than a force shaping professional identity. In Western scholarship, digital change is commonly framed as a response to newsroom transformation, emphasizing speed and competitiveness, while pedagogy remains largely skill-oriented. In Russia, similar developments occur within a distinct institutional context, where journalism education is discussed through modernization and industry alignment (Vartanova, Lukina, 2017; Barabash, Grabelnikov, Gegelova, Osipovskaya et al., 2019). Technological development is presented as a marker of national progress and digital competence (Pantserev, 2021), while classrooms continue to preserve linguistic and ethical tradi tions rooted in a humanistic educational model.

This tension reflects technological solutionism, where tools are expected to address linguistic and educational inequality. Recent studies, particularly in English as a Foreign Language contexts, describe generative AI and machine translation as scaffolding mechanisms that support revision and language management. Wang (2024) reports improved writing outcomes under AI-supported editing, while Kim, Yu, Detrick, Li (2025) show that students view generative AI as a multifunctional assistant for drafting, revision, and feedback. These findings suggest that AI-mediated writing functions as structured support rather than simple substitution.

However, emerging evidence indicates that digital assistance may also replace cognitive effort with imitation. In Russia, where writing is traditionally framed as an ethical and civic practice, this shift carries particular significance. Although multilingual journalism research addresses linguistic adaptation (Kudritskaya, Plastinina, Kushnina, Plekhanova et al., 2024), its implications for motivation and professional dignity remain underexplored.

Large-scale reviews note growing use of AI-assisted writing in language education. Though such tools boost learners’ performance and motivation, research prioritises short-term gains over long-term identity development (Li, Tan, Wang, Lowell, 2025). Recent studies further indicate that automated writing tools reshape language learning practices (Oripova, 2025; Ivanova, Fedorova, Tilga, Artemova, 2025). As efficiency becomes an implicit measure of competence, understandings of authorship shift, redefining expectations of journalistic credibility. Against this background, the present study examines knowledge outsourcing and authorship in Russian journalism education.

The Formative and Ethical Dimensions of Journalism Education in the Age of AI

The formative and ethical dimensions of journalism education have long been framed as a balance between technical skills and intellectual development. Classical scholarship warns that overly instrumental approaches to learning may weaken judgment, responsibility, and public reasoning (Williams, Guglietti, Haney, 2018). This concern is particularly relevant in Russian journalism education, which traditionally combines professional training with a humanistic mission. Earlier discussions of digital production similarly cautioned that an excessive focus on tools may reduce education to operational competence, even as technological adaptation remains important for employability.

Recent research on AI-assisted writing extends this debate by highlighting its dual role. On the one hand, generative AI can function as a form of structured support, assisting with revision, feedback, and complex writing tasks (Wang, 2024; Kim, Yu, Detrick, Li, 2025). On the other hand, large-scale reviews emphasize that such support requires careful pedagogical guidance to avoid replacing reflective learning or weakening intellectual responsibility (Zhou, Li, Chai, Chiu, 2025; Li, Tan, Wang, Lowell, 2025). These perspectives suggest that the key challenge lies in how AI is integrated into professional training rather than in its use alone.

In Russian scholarship, these issues remain closely linked to journalism’s professional mission and civic responsibility (Vartanova, Lukina, 2017). Recent studies focus on modernization and digital transformation (Barabash, Grabelnikov, Gegelova, Osipovskaya et al., 2019), reflecting broader institutional priorities, while questions of authorship and formative learning receive less explicit attention. At the same time, recent work calls for renewed engagement with journalism’s ethical foundations in the post-digital era (Stiekolshchykova, Ziniuk, Melnykova-Kurhanova, Yarova et al., 2025).

Professional dignity, grounded in authorship and accountability, remains a core value in journalism education. However, debates on automation tend to focus more on newsroom practices than on classroom learning. Institutional emphasis on innovation may obscure the formative dimensions of education¹ , while algorithmic systems increasingly shape perceptions of credible writing and responsibility (Parratt-Fernández, Chaparro-Domínguez,

Moreno-Gil, 2024). Against this backdrop, this study examines how journalism education can integrate technological change while sustaining its formative and ethical foundations.

Theoretical Framework

This study interprets the growing practice of knowledge outsourcing among journalism international students in Russia through the combined perspectives of digital capital and digital alienation. Digital capital explains unequal capacities to access and strategically employ AI tools, while digital alienation captures the subjective consequences of delegating cognitive and writing tasks to technological systems. Together, these concepts illuminate both the structural conditions and experiential outcomes of AI-mediated knowledge production in journalism education.

Ragnedda’s research (2018) defines digital capital as a set of social, cultural, and technical resources shaping individuals’ capacity to use and benefit from digital technologies. In the Russian educational context, international students often enter journalism programs with uneven digital capital. They face linguistic barriers, cultural adjustment, and varying levels of digital competence (Ragnedda, Ruiu, Gladkova, 2025). Their reliance on digital systems for writing or translation therefore reflects not only convenience but adaptation to unequal structural conditions (Li, Ling, Wang, Yin, 2026). In this sense, knowledge outsourcing can be understood as an adaptive strategy shaped by differences in digital capital, allowing students to meet institutional expectations while managing linguistic insecurity. However, such adaptation may also limit the accumulation of symbolic and linguistic capital necessary for long-term professional development.

The concept of digital alienation extends this analysis by addressing the experiential dimension of mediated learning. When technol ogical assistance reduces the effort required for expression, students may become distanced from the reflective processes through which meaning develops (Hassan, 2020). Digital alienation therefore describes the gradual separation between learners and the intellectual labour of writing, transforming writing from an exploratory activity into a managed task. Knowledge outsourcing, from this perspective, reflects the cumulative interaction of unequal digital capital and redistributed authorship on students’ engagement with learning. By linking structural inequality, discursive responsibility, and experiential distancing, this framework highlights a central challenge for journalism education in Russia: sustaining reflective authorship within multilingual and technologically mediated environments.

Methodology

This study adopted a qualitative design combining classroom ethnography and semi-structured interviews to explore how journalism international students in Moscow experience the growing tendency of knowledge outsourcing in their writing and learning. Fieldwork was conducted between September 2024 and October 2025 at the Faculty of Journalism, Lomonosov Moscow State University, with supplementary data collected at the Peoples’ Friendship University of Russia. This approach allowed the researcher to observe everyday classroom practices while also capturing students’ personal reflections through in-depth conversations.

Research Design and Sampling

The primary study involved 23 international students representing diverse linguistic and cultural backgrounds and 3 senior lecturers, including participants from China, Vietnam, Indonesia, South Korea, and several African countries. Student participants were selected through purposive sampling based on four criteria: nationality, programme language (English or Russian), stage of study, and enrolment in journalism or communication-related programmes requiring intensive writing, with all students having first languages other than Russian or English. Such diversity allowed the study to capture contrasting patterns of language confidence, workload pressure, and institutional expectations in relation to writing behaviour.

Ten journalism courses were observed across one academic year, generating extensive field notes that documented classroom interaction, task completion, and students’ strategies for managing language-related challenges. Semi-structured interviews, lasting between forty and sixty minutes, were conducted with students to explore their experiences with translation practices, drafting support, and perceptions of authorship and professional development. Three senior lecturers were also interviewed to provide complementary perspectives on teaching expectations and assessment practices. In addition, two expert interviews were conducted with senior academic administrators to provide an institutional perspective on potential educational responses to the issues identified in the primary data.

Data Collection and Researcher Reflexivity

The researcher, being a student at the Faculty of Journalism, had direct access to classrooms and peer communication. To maintain neutrality, a reflexive diary was kept throughout the study to record positional assumptions and interpretive decisions. Observations focused on how students approached journalistic writing tasks, revised or delegated parts of their work, and how teachers responded during workshops and feedback sessions. Interview data were recorded with consent, transcribed verbatim, and anonymised. Codes such as MSU-E5 or PFUR-R3 were used to protect participant identity while maintaining analytic clarity, with E indicating English-teaching students and R Russian-teaching students. Quotations were included only after participants confirmed their accuracy and comfort with inclusion.

Expert Interviews

To add an institutional perspective, the researcher conducted formal interviews with two senior academic administrators from a Faculty of Journalism at a major Russian university. The interviews were semistructured and explored the faculty’s position on AI tools, curriculum adjustments, and evolving assessment practices. Both experts, who have extensive experience in international journalism education, discussed the linguistic and ethical challenges faced by non-native students and reflected on how AI use has reshaped teaching expectations. To preserve anonymity, they are referred to in this study as Expert A and Expert B. With the interviewees’ permission, their views were included in the analysis. These insights complement student narratives by providing a systemic understanding of how journalism education in Russia responds to technological change and the growing role of AI-mediated learning.

Data Analysis

All materials – field notes, transcripts, and classroom records – were analysed thematically following Braun and Clarke’s 6-stage approach (Terry, Hayfield, Clarke, Braun, 2017): familiarisation, coding, theme generation, theme review, naming, and interpretation. The process was iterative rather than mechanical. Initial codes were developed manually and then refined through repeated comparison across data sources until coherent categories emerged. Four main themes were identified: habitual outsourcing, professional identity, authorship dignity, and AI hallucination risks. To enhance analytic reliability, two academic peers independently reviewed a sample of coded transcripts and discussed thematic summaries. Minor differences were reconciled through discussion and revision until conceptual agreement was reached.

Ethical Considerations

To strengthen credibility, data triangulation was applied through classroom observations, interviews, and expert input. Member checking was conducted during follow-up meetings, where participants reviewed and clarified thematic summaries. All participants provided written consent, and confidentiality was strictly maintained through secure data storage and anonymisation of transcripts. Participation was voluntary, with the option to withdraw at any stage. The researcher ensured ethical transparency and aimed to accurately represent participants’ perspectives without overinterpretation.

Findings

Theme 1. From Assistance to Dependence: The Normalisation of AI-Mediated Writing

When the author entered the journalism programme in September 2024, AI tools were already widely used in the classroom, with nearly the entire cohort relying on them in daily writing tasks. At the initial stage, this practice appeared understandable, as international students were adjusting to a new academic environment and working in non-native languages. AI functioned as a supportive assistant, helping students grasp course content and meet writing requirements. However, this reliance did not appear to fade with improved language competence. Even after becoming familiar with the academic context, many continued to use AI as a primary writing aid. What began as temporary support during adaptation gradually developed into a stable writing habit extending to both classroom exercises and assessed coursework.

For students entering international education for the first time, language shame remained a strong emotional driver. Previously, translation tools were commonly used, but the emergence of AI provided a faster way to manage linguistic insecurity.

“I believe Chinese students share a common challenge: while excelling in reading and writing, they often struggle with spoken English. AI’s rapid generation enables me to express my thoughts promptly.” (MSU, E1)

Although universities emphasized independent thinking and academic literacy, students relied on AI largely because peers used it widely and because it offered efficiency and accuracy. Comparisons focused on speed rather than authorship. Those who attempted to write without AI often felt slower and less confident, framing reliance on AI as a rational response rather than an ethical concern.

“I also aspire to become someone who can independently complete academic tasks, but while my classmates finish in ten minutes, it takes me two hours – this frustrates me immensely.” (MSU, R6)

“In class, the instructor asked us to answer questions. Most people found the answers after consulting AI, but I knew nothing – I felt so embarrassed. Since then, I’ve started relying on AI too.” (MSU, E8)

Participants described changes in their writing behaviour associated with frequent AI use. Students spent less time struggling with expression and more time editing AI outputs, while linguistic risk-taking declined. Several students reported selecting suggested wording rather than searching for precise expressions themselves, making writing increasingly a process of selection rather than construction.

Importantly, students described this dependence not as loss but as relief. AI reduced anxiety related to grades and language exposure, yet some participants felt that responsibility became less clearly defined. Errors were attributed to the tool, while originality was less fully claimed as personal, and writing gradually felt less owned.

“AI is smarter than me and more creative than I am. The mistakes made by AI, I will surely make them too.” (PFUR, E18)

In this sense, AI dependence became culturally normalized within the learning environment, reshaping everyday writing practices without formal policies or explicit debate and gradually redefining what it meant to “write” in journalism education.

Theme 2. Changing Perceptions of Writing Identity and Authorship

As reliance on AI-mediated writing became routine, many international students described a shift in how they related to their texts. Writing was less often experienced as a process of thinking through language and more as managing output. Students focused on surface quality rather than meaning, which reduced emotional and intellectual investment.

“AI is more logical than me. The things I write seem to have no value.” (MSU, E10)

For non-native speakers, this shift carried particular significance. Writing had been central to their development as journalists and a source of confidence. With frequent AI use, this struggle became less visible. While writing became faster, some students described it as less personal.

“Sometimes I feel like all the news articles I write have the same vibe, just with different words.” (MSU, E14)

This shift also influenced perceptions of authorship. Many students expressed uncertainty about whether their work could be fully considered their own, though this rarely appeared as explicit ethical conflict. Instead, it emerged as reduced engagement: students were less inclined to revise independently or defend stylistic choices, and lecturer feedback was sometimes experienced as less meaningful.

“Facing the teacher’s praise for my news report, I seemed not to be as happy as before.” (MSU, R3)

In journalism education, authorship is closely linked to professional identity. When writing becomes technologically mediated, some students reported difficulty recognising themselves as active authors, despite producing competent texts. This shift appeared gradual and shaped how students related to writing and professional development.

A frequency overview based on 26 interviews further illustrates these patterns. Most participants (n = 24) associated AI use with efficiency, followed by peer influence and language anxiety (n = 18 each), while study pressure and reduced creativity were mentioned less frequently. Although not statistical, this overview supports the thematic interpretation.

Theme 3. Students’ Perceptions of Language Development

and Career Anxiety

In everyday study life, language was central across lectures, readings, and writing tasks. For journalism international students, it functioned not only as a tool but also as a marker of competence. Russian was necessary for study and daily life, while English was linked to mobility and future employment, shaping how students imagined their professional roles.

As AI tools became integrated into writing practices, discussions about language shifted. Students spoke less about learning vocabulary or structure and more about completing tasks efficiently. Writing appeared faster but often more neutral. During observations, students asked fewer questions about expression and focused more on prompts and templates. When AI was unavailable, hesitation in starting writing became more visible.

“The thing I dread most right now is when the instructor asks me to write on the spot without any preparation.” (MSU, R12)

This tension was also reflected in internship-related experiences. Students who had worked at Russia Today and other media outlets noted that language ability remained a key factor in recruitment, particularly the ability to work across languages independently.

“My HR manager mentioned that some positions favor international students due to their language proficiency. However, since the advent of AI, an increasing number of international students have demonstrated poor language organization skills without AI.” (MSU, R16)

“AI can produce news quickly and efficiently. I fear being replaced by AI in the future. I even think this is no longer an advantage for journalism students – students from any major could become journalists someday.” (PFUR, R20)

Several students expressed concern that frequent AI use might weaken this advantage. While AI improved fluency, some participants felt that it reduced opportunities to practice language independently.

Career-related concerns often followed. Students suggested that these changes might not be immediately visible but could emerge in professional contexts such as interviews or newsroom work. In this sense, AI use appeared to shape both learning practices and expectations about future careers.

Theme 4. AI Hallucinations and the Limits of Students’ Critical Judgment

From the lecturers’ perspective, concerns about AI use extended beyond students’ writing habits. Several instructors noted that many international students showed limited familiarity with basic AI literacy practices, often treating AI-generated content as ready-made knowledge rather than material requiring verification. This was seen as particularly problematic in journalism education, where accuracy and source awareness are essential.

These risks were more evident in topics related to international affairs and political history. Lecturers observed that AI-generated responses could appear neutral and academic while lacking contextual depth, leading some students to accept simplified interpretations that were not always immediately detectable in written work.

“The use of AI to generate geopolitical knowledge remains controversial. One student labeled certain countries as authoritarian without carefully examining the evidence.” (MSU Senior Lecturer, 1)

During assessment, lecturers reported typical signs associated with AI hallucination. Some students cited unverifiable sources or used unfamiliar concepts not covered in class. When asked to explain these references, several appeared confident but demonstrated only partial understanding.

AI-generated texts were often described as fluent and well structured, creating an impression of competence. However, during seminars, some students struggled to explain ideas in their own words or justify their arguments, suggesting a gap between textual quality and conceptual understanding.

“AI illusions can lead students to overestimate their academic abilities. At the same time, they complicate our assessment of learning outcomes.” (MSU Senior Lecturer, 3)

Overall, these patterns suggest that AI use may complicate both students’ learning processes and lecturers’ evaluation of knowledge, particularly when surface-level correctness obscures deeper understanding.

Once information saturation was reached and four thematic frameworks were identified, an interpretive model of AI adaptive reliance in journalism education was developed. The model outlines analytically inferred patterns associated with international students’ engagement in AI-driven knowledge outsourcing, illustrating potential short- and longer-term influences observed across the data rather than fixed cause-and-effect outcomes.

Expert Perspectives: Pedagogical Responsibility in the Age of AI

In the interviews with Expert A and Expert B, both senior academic administrators of the Faculty of Journalism at a major university, a shared conviction emerged: journalism education must neither resist nor romanticize artificial intelligence but reinterpret its moral and formative role within higher education. Their reflections reveal a paradigm shift from technical adaptation toward the cultivation of critical, ethically grounded, and self-reflective AI literacy.

Expert A underscored that AI literacy is inextricably linked to digital capital. Digital capital once confined to basic technological competence, but now demands a redefinition that includes AI literacy as moral literacy. In their view, the rapid expansion of AI in both academic and professional journalism necessitates that students develop a new form of “AI capital,” encompassing not only functional knowledge but also ethical discernment and epistemic caution. “Everybody today should have a certain level of digital capital,” they noted, “but the higher the level, the more important becomes one’s awareness of how AI shapes our thinking and our choices.” In their view, AI literacy must be inseparable from the intellectual discipline of verification, the same habit that underpins journalistic credibility.

This ethical imperative extends to pedagogy. Expert A warned that while AI can assist with language polishing, structural editing, or stylistic improvement, it should never replace human authorship. “Students must learn to be critical of what they receive from ChatGPT or DeepSeek,” they emphasized, “because information that sounds authoritative may still be wrong.” Their remarks illuminate an emerging institutional consensus: AI may be tolerated as a linguistic tool, but not as an epistemic substitute. Reflecting this stance, a media journal, where they serve as an editorial board member, has recently adopted a clear AI policy – permitting its use for grammar correction but prohibiting content generation. Such distinctions exemplify how journalism faculties are beginning to regulate not only AI usage but also its ethical boundaries.

Expert B’s perspective complements this moral framing with a focus on curricular transformation. They argued that AI competence must be institutionalized within journalism education rather than left to informal experimentation. “It is always better to make AI our friend, not our enemy,” they observed, urging that educators themselves become AIliterate to guide students effectively. Without structured AI training, curricula risk becoming obsolete and misaligned with the labor market. Yet Expert B cautioned against equating modernization with automation: “What students need is not technical imitation, but intelligent application – learning how AI can strengthen, not replace, their analytical and creative capacities.”

Both experts emphasized the centrality of critical thinking as the intellectual bridge between technological fluency and professional dignity. Expert A described critical reasoning as “the last line of defense that differentiates human journalists from machines.” Similarly, Expert B linked critical thinking to employability: graduates capable of reflecting on how AI mediates their creative process will be better positioned to navigate uncertain professional futures. In this sense, AI literacy becomes not only a skill but a moral posture – a renewed commitment to intellectual independence within algorithmic environments.

Institutionally, some faculties of journalism have begun experimenting with integrative strategies that respond to these challenges. Expert B highlighted initiatives such as “Industrial Week,” during which media professionals engage directly with students to discuss how automation reshapes newsroom practices. These encounters, they explained, “restore journalism’s human dimension by showing students that creativity and ethical judgment remain irreplaceable.” Such efforts align with Expert A’s conviction that education should cultivate discernment rather than dependency.

Together, the two experts articulate a vision of journalism education that reclaims its formative mission amid technological transformation. Rather than viewing AI as an external threat, they interpret it as a mirror reflecting journalism’s enduring ethical foundations – truthfulness, authorship, and responsibility. Their insights suggest that the future of journalism pedagogy in Russia will depend less on mastering new tools than on redefining what it means to think, write, and act humanly in a mediated world.

Discussion and Limitations

This discussion moves beyond description to interpret how AImediated writing becomes normalized in journalism education for international students in Russia. Rather than viewing AI knowledge outsourcing solely as a response to linguistic difficulty or individual strategy, the findings suggest it functions as an adaptation to institutional contexts where fluency, accuracy, and timely completion operate as key indicators of academic performance. This interpretation aligns with studies describing AI as a scaffold supporting multilingual writing (Wang, 2024; Kim, Yu, Detrick, Li, 2025), yet extends prior work by showing how sustained reliance may gradually reposition writing from meaning construction toward a more procedural and efficiency-oriented task.

These changes have implications for authorship. While students remain formally responsible for their texts, the process of reasoning through language becomes partially mediated by technological systems. Similar to concerns raised in discussions of instrumental learning in journalism education (Williams, Guglietti, Haney, 2018), authorship may shift from an embodied practice of writing toward a more supervisory role involving selection, coordination, and verification. Responsibility is therefore not removed but reconfigured in ways that are less visible in everyday learning processes, raising questions about how professional accountability develops within AIsupported environments.

From a theoretical perspective, this study contributes by framing AI use as an educational mechanism rather than only a technological or ethical issue. The concept of AI knowledge outsourcing highlights how institutional expectations, linguistic demands, and technological affordances interact to reshape writing practices and redistribute cognitive effort. By situating these findings alongside existing debates on digital literacy and mediated learning (Zhou, Li, Chai, Chiu, 2025; Li, Tan, Wang, Lowell, 2025), the study extends current discussions of multilingual journalism education through an emphasis on authorship, responsibility, and long-term professional formation.

Several limitations should be acknowledged. The study is based on a small, context-specific sample of journalism international students from two leading Moscow universities, which limits the broader applicability of the findings to other regions or institutional settings in Russia. The qualitative design, relying on classroom observation and interviews, captures participants’ experiences rather than measurable outcomes. In addition, the ethnographic approach is subject to “technical blindness,” as it cannot fully account for all students’ digital interactions with AI tools during writing processes. The study also focuses on a specific disciplinary context and does not systematically address wider institutional or policy-level factors. Therefore, the findings should be understood as exploratory, aimed at problematising AI-mediated writing practices rather than offering generalisable conclusions. Future research could adopt comparative and multi-site approaches to examine these dynamics across diverse educational contexts.

Conclusion

This study addresses the first research question by examining the experiences and concerns of journalism international students in Moscow regarding AI-mediated knowledge outsourcing, as well as the underlying factors contributing to this practice. The findings suggest that reliance on AI-mediated writing is shaped not only by workload pressure and linguistic insecurity, but also by institutional expectations that reward speed and formal accuracy. What initially appeared as a practical response to multilingual academic demands gradually developed into patterns of delegated writing. This observation partly aligns with studies describing AI as a supportive scaffold (Wang, 2024; Kim, Yu, Detrick, Li, 2025), while extending prior research by noting that sustained dependence calls for careful pedagogical guidance to preserve reflective learning, and is associated with changes in how students describe their writing engagement (Zhou, Li, Chai, Chiu, 2025; Li, Tan, Wang, Lowell, 2025).

The second research question concerns how journalism education might respond to this challenge. Interviews with two experts indicate that pedagogical reflection in Russian journalism education increasingly focuses on balancing technological integration with intellectual discipline. Rather than rejecting digital tools, instructors emphasised maintaining writing as a site of reasoning, verification, and responsibility. This aligns with earlier literature warning against reducing professional training to operational competence (Williams, Guglietti, Haney, 2018)¹ , while adding qualitative insight into multilingual classroom contexts.

Taken together, the findings suggest that digital literacy in journalism education is a multidimensional competence encompassing technical, linguistic, and ethical aspects. While previous studies recognise the supportive role of AI tools, the present study indicates that their longerterm influence depends on instructional design and assessment practices. Alternative interpretations remain possible, as patterns described here may reflect adaptive strategies rather than stable skill decline.

In practical terms, the study highlights the need for curriculum models that support reflective writing while acknowledging mediated learning environments. Journalism education thus faces not only a technological challenge, but an educational one: how to sustain critical judgment, language development, and professional accountability.

Institutional Sensitivity Note

This study primarily involves students from Lomonosov Moscow State University and the Peoples’ Friendship University of Russia, both recognized for their engagement in international and multilingual journalism education. The research does not evaluate either institution. Based on fully anonymized data, it examines students’ adaptive learning in AI-mediated environments and aims to inform pedagogy, curriculum development, digital literacy, ethics, and professional formation.

Примечания

¹ Demmar K., Neff T. (2023) Generative AI in journalism education: Mapping the state of an emerging space of concerns, opportunities, and strategies. University of Leicester. Available at: https://hdl.handle.net/2381/24648093.v1 (accessed: 23.06.2026).

Библиография

Barabash V. V., Grabelnikov A. A., Gegelova N. S., Osipovskaya E. A., Rebrova A. D. (2019) Challenges for modernization of journalism education in Russian universities. In Education Environment for the Information Age. European Proceedings of Social and Behavioural Sciences. Vol. 69. Future Academy. Pp. 117–124. DOI: 10.15405/epsbs.2019.09.02.14

Hassan R. (2020) Digital alienation. In The Condition of Digitality: A Post-Modern Marxism for the Practice of Digital Life. London: University of Westminster Press, Pp. 159–189. DOI: 10.16997/book44.g

Ivanova A., Fedorova N., Tilga S., Artemova E. (2025) Voices of Freelance Professional Writers on AI: Limitations, Expectations, and Fears. arXiv preprint arXiv:2504.05008

Kim J., Yu S., Detrick R., Li N. (2025) Exploring students‘ perspectives on generative AI-assisted academic writing. Education and Information Technologies 30 (1): 1265–1300. DOI: 10.1007/s10639-024-12878-7

Kudritskaya M., Plastinina N., Kushnina L., Plekhanova Y., Matytcina M., Stepanova M. (2024) Balancing innovation with ethics: AI applications for enhancing language competence in academic writing and reading. In 2024 4th International Conference on Technology Enhanced Learning in Higher Education (TELE). IEEE. Pp. 380–385. DOI: 10.1109/tele62556.2024.10605668

Li B., Tan Y. L., Wang C., Lowell V. (2025) Two years of innovation: A systematic review of empirical generative AI research in language learning and teaching. Computers and Education: Artificial Intelligence 9: 100445. DOI: 10.1016/j.caeai.2025.100445

Li Y., Ling Y., Wang Z., Yin H. (2026) Digital divide, digital capital, and educational dropout risk: structural and psychological pathways of digital inequality. In Proceedings of the 2026 International Conference on Big Data and Informatization Education. ACM. Pp. 771–776. DOI: 10.1145/3806980.3807102

Murafa S. V., Hoc H. T., Jingyu X. (2025) Cross-cultural study of the potential applications of artificial intelligence in higher education (Russia, Vietnam, and China). In Integration Strategies of Generative AI in Higher Education. IGI Global. Pp. 229–246. DOI: 10.4018/979-8-3693-5518-3.ch011

Oripova N. (2025) AI and its influence on English language learning, communicative competence and journalism education. Mezhdunarodnyi mul’tidistsiplinarnyi zhurnal issledovanii i razrabotok 12 (3): 377–381.

Pantserev K. (2021) Media education in the age of AI and smart media. Mediaobrazovanie 17 (3): 507–514. DOI: 10.13187/me.2021.3.507

Parratt-Fernández S., Chaparro-Domínguez M.-Á., Moreno-Gil V. (2024) Journalistic AI codes of ethics: Analyzing academia‘s contributions to their development and improvement. Profesional de la información 33 (6): e330602. DOI: 10.3145/epi.2024.0602

Ragnedda M. (2018) Conceptualizing digital capital. Telematics and Informatics 35 (8): 2366–2375. DOI: 10.1016/j.tele.2018.10.006

Ragnedda M., Ruiu M. L., Gladkova A. (2025) Reframing digital capital in unequal contexts. World of Media. Journal of Russian Media and Journalism Studies 3: 5–29.

Shesterkina L., Krasavina A. (2021) Journalism education in the era of social networks and artificial intelligence: Digital technologies and ethical values. Proceedings of International Online Round Table.

Stiekolshchykova V., Ziniuk A., Melnykova-Kurhanova O., Yarova A., Lobanov D., Soldatenko I. (2025) Personal factors shaping journalists‘ professionalism in today‘s military journalism. Revista Romaneasca pentru Educatie Multidimensionala 17 (2): 629–651. DOI: 10.18662/rrem/17.2/998

Terry G., Hayfield N., Clarke V., Braun V. (2017) Thematic analysis. In The SAGE Handbook of Qualitative Research in Psychology. London: SAGE. Pp. 17–37. DOI: 10.4135/9781526405555.n2

Vartanova E., Lukina M. (2017) Russian journalism education: Challenging media change and educational reform. Journalism & Mass Communication Educator 72 (3): 274–284. DOI: 10.1177/1077695817719137

Wang Y. (2024) Cognitive and sociocultural dynamics of self-regulated use of machine translation and generative AI tools in academic EFL writing. System 126: 103505. DOI: 10.1016/j.system.2024.103505

Williams A., Guglietti M. V., Haney S. (2018) Journalism students‘ professional identity in the making: Implications for education and practice. Journalism 19 (6): 820–836. DOI: 10.1177/1464884917692344

Yan L., Greiff S., Teuber Z., Gasevic D. (2024) Promises and challenges of generative artificial intelligence for human learning. Nature Human Behaviour 8 (10): 1839–1850. DOI: 10.1038/s41562-024-02004-5

Zhou X., Li Y., Chai C. S., Chiu T. K. F. (2025) Defining, enhancing, and assessing artificial intelligence literacy and competency in K-12 education from a systematic review. Interactive Learning Environments 33 (10): 5766–5788. DOI: 10.1080/10494820.2025.2487538

Как цитировать: Ли Я. ИИ-аутсорсинг знаний и профессиональное становление иностранных студентов-журналистов в московских университетах // Вестник Московского университета. Серия 10. Журналистика. 2026. № 3. С. 137–159. DOI: 10.55959/msu.vestnik.journ.3.2026.137159

Поступила в редакцию 29.12.2025