621 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 621 • let. 62, 3/2025 07 Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU* AI IN THE NEWSROOM: CHANGES AND CHALLENGES IN ALBANIA** Abstract. The article explores the implications held by artificial intelli- gence (AI) for newsrooms in Albania through a qualitative research ap- proach. Utilising 24 semi-structured, in-depth interviews with journalists and editors, media professionals’ experiences and perceptions regarding the integration of AI into their practices are studied. The results indicate that AI is employed for information gathering, translation, image verific- ation, and news trend analysis. While identified advantages of AI include time optimisation and improved news quality, challenges persist, such as concerns about information accuracy and scepticism of technology. Journ- alists stress the need for manual verification and the irreplaceable role of the human factor in information security. The use of AI data to identify audience preferences also demonstrates an increase in the capacity of AI to deliver content that resonates with audiences. It is argued in the study that collaboration between AI and journalists is essential to ensure quality and accuracy in journalism. Keywords: Artificial Intelligence, AI, journalism, media, social media, news, newsroom. INTRODUCTION In journalism, an inhuman presence has already entered the realm of news writing. Artificial intelligence (AI) has found applications in newsrooms for news reporting (Biswal and Gouda 2020), newsgathering, and distribution (Deuze and Beckett 2022). Notably, The Washington Post frequently indicates that “it was ‘powered by Heliograf’, which is The Post’s artificial intelligence system” (Tandoc et al. 2020, 549). This shift facilitates the automated production of thousands of news articles daily across various countries (Barrolleta and Sandoval-Martín 2024). For instance, the Associated Press collaborates with startups to introduce * Ramadan Çipuri, PhD, Associate Professor, University of Tirana, Albania; Erlis Çela, PhD, Associate Professor, Bedër University College, Tirana, Albania; Hasan Saliu, PhD, Associate Professor, AAB College, Pristina, Kosovo, e-mail: hasan.saliu@aab-edu.net. ** Research article. DOI: 10.51936/tip.62.3.621 622 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 622 TEORIJA IN PRAKSA AI innovations in news production, while the Chinese government is urging its state-run Xinhua news agency to integrate AI to create a high-tech newsroom (Zhang et al. 2023). These developments highlight the evolving dynamics within media and communication landscapes (Saliu 2024) that are transforming news gathering, production, and distribution processes (Pavlik 2023). The increased use of AI has also raised ethical concerns and prompted new professional practices in journalism (Deuze and Beckett 2022; Al-Zoubi et al. 2024). These advancements are leading to radical changes in newsroom prac- tices that bring into question journalists’ profiles and the professional knowledge needed for information collection and distribution. These significant transform - ations – part of the over 500 years of journalistic tradition – have called into question the Zoabi compatibility of such integrated automation with traditional journalism values and its impact on perceptions of credibility (Tandoc et al. 2020). Concerns about trust in news have thus emerged alongside the adoption of AI and automated journalism (Ali and Hassoun 2019) given that producing text and audiovisual content has become easier than ever, potentially involving biased or misleading outputs (Illia et al. 2023). Moreover, it is not only Western countries that are adopting AI. When it comes to the utilisation of new technologies (including the Internet and social media), developed and developing countries exhibit minimal differences. Annual reports from the Reuters Institute show the inhabitants of least developed coun- tries are often more informed online than their counterparts in developed coun- tries are, who tend to prefer offline media for information (Newman et al. 2024; 2022; 2021). In Albania, where the consumption of online media is widespread, the incorporation of AI in journalism offers a distinct chance to analyse the motives, attitudes and changes in work practices caused by AI integration. To understand how AI is adopted and perceived in journalism in Albania, it is essential to consider the national media circumstances. Albania’s media is char- acterised by high ownership concentration, political affiliations among main- stream media organisations, and a relatively developed digital infrastructure. While the number of media channels suggests pluralism, concerns regarding editorial independence, transparency of funding, and political pressure persist (Media Ownership Monitor Albania 2023; European Commission 2023). These structural factors have a major impact on how journalists operate, interpret, apply or resist AI technologies in their day-to-day work. As Cools and Diako- poulos (2024) emphasise, the adoption of AI in newsrooms cannot be separated from local political and institutional contexts, which influence both the poten- tial and limitations of technology. This article aims to investigate the role of AI in Albanian media outlets, stressing its applications, perceived advantages, and problems, together with the larger implications of AI for journalistic practices. Hence, the research questions are: 623 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 623 • let. 62, 3/2025 • For which purposes do media journalists in Albania use AI for? • How do Albanian journalists and editors perceive the influence of AI on the quality, accuracy and credibility of news content? • How does AI help with identifying audience preferences? LITERATURE REVIEW The emergence of automated news writing has added to the decline in the perceived importance of journalists in the newsroom since this process requires minimal human intervention after the initial programming. This shift speeds up production and alleviates newsrooms’ budget constraints (Graefe 2016). How- ever, non-human actors are not limited to news writing. In a study of AI use in newsrooms, Beckett and Yaseen (2023, 4) analysed 105 news and media organisa- tions across 46 countries and found that AI is employed for newsgathering, news production, and distribution, encompassing all stages of content creation “from ideation to publishing”. More than half the respondents from these media organ- isations mentioned increased efficiency and productivity as the core objectives driving AI adoption. Journalists and media executives seek to automate mono- tonous tasks, thereby streamlining workflows and allowing journalists to engage in more creative, relevant and innovative work (Beckett and Yaseen 2023). The integration of AI into news collection, report design, and distribution has transformed news production practices (Munoriyarwa et al. 2023), a trend that began with technological advancements and the rise of social media, altering the evaluation models and criteria for media systems (Houston 2023; Saliu et al. 2023). Although news was traditionally produced in-house, the use of freelancers has become common, enabling journalists to work outside the newsroom in a networked manner (Deuze and Witschge 2019). The growth of online media has facilitated this shift, prompting a rise in media entrepreneurs. Journalistic man- agers now face the challenge of balancing editorial and professional responsib- ilities as the traditional definition of journalism evolves (Deuze and Witschge 2019; Hölsgens et al. 2020). The advent of AI means that today journalism can occur even if human journalists are absent. These changes are closely tied to the need to balance the speed and urgency demanded by online media these days with the need for reliability, objectivity and accuracy in journalism, all while meeting public expectations for impactful and engaging reporting (Srisaracam 2023). AI is revolutionising journalism by transforming how news is gathered, edited and presented. Media organisations nowadays produce a substantial volume of news content that is collected, edited and presented with a high degree of professionalism (Al-Zoubi et al. 2024). In the realm of news and journalism, AI is often associated with various uses of AI-as-a-Service (AIaaS), which provides organisations with access to specific AI capabilities through cloud computing. This can range from chatbots and know- ledge maps to computer vision and voice recognition (Simon 2024). 624 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 624 TEORIJA IN PRAKSA From a managerial perspective, automated journalism is justified by rational business considerations. AI can facilitate numerous tasks in every stage of the reporting process, from gathering to producing, verifying and distributing news (Cools and Diakopoulos 2024). Journalists can automate various aspects of news production, such as data discovery and verification, graphic production, pub- lishing with selected filters, and automatic article tagging (Lopez et al. 2023). AI-supported journalism enables numerous articles to be produced rapidly, even in multiple languages, considerably reducing production costs and enhancing profit margins (Karim and Porlezza 2020). AI is also making significant strides in the areas of news gathering and verification. Journalism is fundamentally a verification discipline (Kovach and Rosenstiel 2021), and the need for verification has become ever more critical in an era when misinformation can spread quickly (Tariq et al. 2024). Even though the term fake news gained prominence in 2016, it refers to the long-standing issue of misinformation (Barclay 2018). Disinformation, misinformation and propa- ganda all have historical precedents, but the current environment is uncharted territory. The Internet and social media allow individuals to disseminate their narratives, some of which might be untrue (Saliu 2023; Saliu and Llunji 2022). As defined by Allcott and Gentzkow (2017, 213), “Fake news is intentionally and verifiably false and could mislead readers”. This underscores the significance held by AI today given the vast volume of news circulating in the online sphere. Journalists can utilise AI-powered tools to identify patterns and trends in large datasets, as well as to validate sources and information. AI is increasingly relied on to track trending topics and breaking news on social media platforms, permitting journalists to cover the most press- ing stories of the day (Diakopoulos and Koliska 2016). This process has become considerably easier today, particularly with the use of ChatGPT, currently the most widely used generative AI software system in the world (Samo and Svetlik 2024). Digital transformations have seen new forms of journalism emerge that chal- lenge epistemological and professional themes (Cools et al. 2022). As technolo- gical adoption rises, so too do ethical concerns in journalism (Deuze and Beckett 2022), especially as regards the impact of AI algorithms on journalistic processes. In a survey of 105 news organisations in 46 countries conducted by Beckett and Yaseen (2023), 60% of respondents expressed concerns about the integration of AI into journalism. Such concerns largely arise because algorithms can intro- duce biases in content evaluation and may produce information whose accuracy and reliability may be challenged (Cools and Diakopoulos 2024). Further, the mysterious nature of certain algorithms complicates transparency and human decision-making (Beckett and Yaseen 2023). These changes are primarily concerned with balancing the speed and urgency required by online media these days with the need for reliability, objectivity and accuracy in journalism, while also addressing business profitability and public 625 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 625 • let. 62, 3/2025 expectations for impactful reporting (Srisaracam 2023). The change in focus to audience preferences has fundamentally altered journalism. Catering to pub- lic interests calls for providing the information that the public desires rather than what journalism professionals believe the public should be informed about (Kovach and Rosenstiel 2021; Saliu et al. 2024). In digital journalism, newsrooms have implemented audience analytics to identify the most-clicked and read art- icles, as well as to measure the time spent on particular news items, thereby assessing audience consumption patterns (Lee and Tandoc 2017). However, the influence of audience metrics on news production remains a critical con- cern (Dodds et al. 2023). “Audience metrics are quantitative information used to monitor, measure, and analyse various aspects of the audience, including, but not limited to, the size and composition of the audience, audience behaviour, and audience engagement” (Wang 2017, 472). The implementation of audience ana- lytics in newsrooms acts as external pressure exerted by web analytics compan- ies, which “promote profit-making orientations in the newsroom that challenge and influence the professional values and norms of news production” (Lamot and Paulussen 2020, 360). Many newsrooms utilise audience metrics not simply to align media strategies with audience interests but also to evaluate journalists’ performance (Lee and Tandoc 2017). Such dynamics resonate with the ideas presented by John Pavlik (2000) who noted that journalism has always been shaped by technology. He argued that technological changes impact: (1) the ways journalists conduct their work; (2) the nature of news content; (3) the structure and organisation of both newsrooms and the news industry; and (4) the nature of relationships within and between news organisations, journalists, and their diverse publics, including audiences, competitors, news sources, sponsors, and regulatory entities (Pavlik 2000, 229). Some scholars suggest that journalists engage in a continuous negotiation pro- cess when confronted with technological innovations, initially employing new tools in accordance with traditional routines, but eventually developing new practices around these technologies to balance their adherence to established norms with the need to adapt to external market pressures in news production (Nelson and Tandoc 2018; Tandoc 2014; Tandoc and Maitra 2017). In the con- temporary hypermedia landscape, establishing a balanced dynamic among the audience, responsible political actors as carriers of messages, and the media is vital for promoting transparent communication and supporting the integrity of democratic discourse (Verčič 2022). METHODOLOGY The presented study employs a qualitative research approach to explore the implications held by AI for newsroom organisations in Albania. The choice of qualitative methodology reflects the need to understand the experiences, percep- tions and practices of journalists and editors as they increasingly incorporate AI into their professional routines. Qualitative research is particularly well suited 626 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 626 TEORIJA IN PRAKSA for studying complex social phenomena where meanings and experiences vary across contexts, as demonstrated in journalism studies focused on technology adoption. The primary data for this study were gathered through 24 semi-struc- tured, in-depth interviews with journalists and editors working in various media outlets across Albania. The interview cohort comprises nine editors, including an editor-in-chief and an information director, as well as two participants from online portals and seven from television. The group of interviewed journalists includes fifteen individuals, with seven from online portals, and eight from tele- vision. The study relies on qualitative in-depth, semi-structured interviews (Denzin and Lincoln 2017). Semi-structured interviews are particularly effective for gath- ering consistent information from all participants while allowing them to elab- orate on their thoughts (Pajo 2018). These interviews aimed to collect detailed data regarding media professionals’ experiences within their editorial practices. Numerous studies have identified in-depth interviews as an effective qualitative method for exploring the impact of new technologies on journalism (Jamil 2020). The journalists and newsroom editors considered in this study were chosen based on their roles in the newsroom and their experience with AI tools and applications in journalism. The sample includes a mix of editors and journal- ists from various media formats, including audiovisual and online platforms, to capture a broad range of experiences. Media outlets were purposively selected to include leading national newsrooms known for their influence on public dis- course. Although not statistically representative, the sample reflects structural and editorial diversity, consistent with the goals of qualitative research to ensure contextual depth and relevance. Previous studies show the importance of incor- porating multiple perspectives from within-newsroom hierarchies to understand how AI impacts different roles (Dalen 2024). In addition, priority while selecting interviewees was given to those who had directly engaged with AI technologies for automated content generation, news verification, and audience data analysis. The questionnaire employed in the interviews was structured around five main topics: (1) the use of AI in information gathering; (2) news trends and the nature of content; (3) the structure and organisation of newsrooms under the influence of AI; (4) the relationship with audiences; and (5) the advantages and challenges of AI in journalism. Overall, the questions aimed to explore journ- alists’ views on the role, benefits and challenges of AI for journalistic practices. In preparing and conducting the interviews, the methodological approach utilised in prior studies was adapted. The researchers Cools and Diakopoulos (2024) investigated journalists’ perceptions in the Netherlands and Denmark concerning the opportunities and risks associated with AI in newsrooms, employing a qualitative approach grounded in semi-structured interviews. Sim- ilarly, another study used semi-structured interviews to examine the integra- tion of AI into editorial routines, focusing on cases from the BBC and The Times (Lopez et al. 2023). 627 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 627 • let. 62, 3/2025 The interviews were conducted by the authors either face-to-face or via email depending on the participants’ preferences and accessibility. All interviews were conducted in line with established ethical standards for qualitative research. The interviewees provided informed consent before par- ticipating, and their identities were anonymised to assure confidentiality. The research team securely stores the interview transcripts, which are not publicly accessible, in compliance with data protection regulations such as the GDPR (European Commission 2021). Nevertheless, the study strictly adhered to ethical research principles, including informed consent, data anonymisation, and con- fidentiality protocols. Data Analysis Thematic analysis as described by Braun and Clarke (2006) was employed to analyse the interview data. This method offers a systematic approach to identify - ing, analysing and reporting patterns (themes) within qualitative data, making it suitable for examining diverse experiences and perceptions regarding the integ- ration of AI into journalism (Braun and Clarke 2006). The analysis process encompassed several steps, beginning with becoming familiar with the collected data by reading the interview transcripts, then identi- fying and labelling segments of responses pertinent to the research questions. Findings were grouped into themes such as “uses and advantages”, “quality, accuracy and verification”, “trends and identification of audience preferences”, and “challenges and expectations”. The analysis then proceeded with a detailed examination of each theme, using direct quotes from interviewees to illustrate key findings. When processing responses, answers were first categorised by topics and then coded to ensure anonymity as provided for in the questionnaire. Responses were given the codes “Ed.” for editors and “Journalist” for journalists. The codes were further divided into two categories: interviewees employed in television, and those engaged with online portals. The coding format followed this struc- ture: “TV Ed.” + ‘number’ for television editors, and “Portal Ed.” + ‘number’ for portal editors. Television journalists were labelled “TV Journalist” + ‘num- ber’, while those from online portals were designated “Dig. Journalist” + ‘num- ber’. For example, the first editor interviewed from television was coded as “Tv Ed.1”, whereas the response from a journalist affiliated with an online portal was marked “Dig. Journalist2”. RESULTS AI is employed in Albania’s media establishments for information gathering, translation, image verification, and news trend analysis, facilitating the extrac- tion of accurate data. Journalists exhibit varying attitudes to AI; some rely on traditional methods and refrain from using AI to gather political information. Conversely, some portal journalists utilise AI to translate materials and analyse 628 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 628 TEORIJA IN PRAKSA complex data, albeit they always verify the accuracy of the information. Notably, the use of AI in online journalism, particularly on portals, has become preval- ent as these platforms seek to better understand audience preferences. AI is also employed for text-to-podcast conversions, although challenges persist among more experienced journalists and scepticism remains among editors who favour traditional models. Use and Advantages An interviewed journalist summarised how AI is applied in their work as follows: I use AI in cases where I need to translate video material from a language I don’t speak, and to find information about a specific individual, which I then verify with the original article or other sources. I also use it to research a particular event or individual and to analyse large amounts of data. (Dig. Journalist6 2024) Analysis of the interview responses reveals that AI is more widely utilised in portal editorial offices for automating certain information-gathering processes. One journalist noted that news writing itself is becoming more automated, rapid and straightforward: “We are using automation to write simple news, such as sports score reports or statistical information, which allows us to focus on more in-depth analysis” (Dig. Journalist1 2024). The same journalist added, “Another aspect is the transcription of interviews. Journalists no longer handle the tran- scription themselves; instead, they rely on AI-generated applications” (Dig. Journalist1 2024). Even in its initial stages, the use of AI is seen to offer several advantages. The most frequently mentioned benefits include time optimisation and the poten- tial to enhance news quality. One journalist encapsulates these advantages: “With AI, we optimise the time spent searching for and collecting information; we summarise lengthy texts into key points; we compile different news; and we transcribe statements swiftly in both Albanian and foreign languages” (TV Ed.8 2024). Many journalists view AI more as a preliminary tool for generating ideas rather than as a direct influence on the final content (Dig. Journalist4 2024; TV Journalist8 2024). A portal journalist stressed that, given the current inclination for video over reading, AI facilitates the conversion of text into audio: The Voice Maker app is a program that transforms the entire text of news articles into podcast versions. Its impact has been positive as feedback from our audience indicates it allows them to listen to the news while driving or engaging in other activities. The audio-podcast format has made the content more enjoyable and minimised the audience’s time commitment. (Dig. Journalist1 2024) 629 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 629 • let. 62, 3/2025 Others mentioned that AI assists in video creation and editing, as well as transcribing interviews, thus saving time and allowing journalists to concentrate on more creative aspects. “ AI offers superior translations compared to traditional platforms like Google Translate, especially ChatGPT, which also edits the text, particularly in English” (Dig. Journalist6 2024). Overall, the chief advantages include time savings, increased accuracy, and improved data analysis and audi- ence behaviour insights. However, some journalists express concerns with AI’s emphasis on speed in information gathering. They contend that these technologies prioritise expedi- ence, necessitating manual verification. Some television newsroom editors have not fully integrated AI, noting they are still in the exploratory phase to determ- ine its optimal applications. They assert that the proper implementation of AI will require time. Quality, Accuracy and Verification The Albanian media’s incorporation of AI has influenced the quality and accuracy of information, albeit attitudes among journalists on this matter vary. Television editors appreciate AI tools for enhancing accuracy, aiding with source verification, and filtering out misinformation. One interviewee argued, “AI is what distinguishes news from rumours”. He further explained, “In an era when false information creates chaos, robust AI tools help us produce high-quality news based solely on verified facts (sources)” (TV Ed.1 2024). This suggests that AI can help significantly with improving news quality. Still, he pointed out that, despite AI’s role in expediting data processing, challenges with accuracy remain, particularly concerning current information, as AI platforms often rely on out- dated data. Conversely, some portal journalists acknowledge the efficiency that AI brings, but point to the need to filter information to ensure its accuracy. From the per- spective of television journalists, scepticism exists regarding AI’s accuracy. Some maintain that they still depend on manual research and traditional methods, reflecting a level of mistrust in new technologies, notably AI. Nonetheless, others concur that AI speeds up data collection, asserting that its integration into these processes should always be accompanied by careful source verification. According to the interviewed journalists and editors, the human factor remains critical in the Albanian media for verifying the information collected by AI. A significant number of editors and journalists viewed AI as a valuable tool for data collection and analysis, yet they underscored that the role of journalists is indispensable to guarantee the quality and accuracy of news. One TV editor argued that while AI can provide the initial data analysis, the human journalist’s role is irreplaceable: “AI can collect and analyse data, but it is journalists who assign the proper meaning and narrative. Accuracy hinges on the collaboration between humans and technology, each fulfilling an irreplaceable role” (TV Ed.1 2024). 630 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 630 TEORIJA IN PRAKSA The division of responsibilities between AI and journalists is a necessary col- laboration to assure a high level of journalistic quality. “Information generated by AI platforms must be verified by journalists to ensure the complete accur- acy of the information and the context in which it will be utilised” (Portal Ed.1 2024). The interviewees emphasised that AI cannot operate without human inter- vention, particularly with fact-checking and contextualisation. They stress the journalists’ role as a crucial filter to make sure of the reliability and accuracy of information. “Journalists and editors often serve as filters, reviewing and inter- preting the data provided by AI, especially in cases where more in-depth analysis is required” (TV Ed.8 2024). Nevertheless, some journalists and editors (both on television and with online portals) maintained scepticism over AI’s capabilities in this respect. They argued that AI cannot detect, analyse or verify real-time events, highlighting that this technology relies solely on historical data and cannot provide direct insights into current occurrences. One interviewee expressed such scepticism, stating, “How can AI detect or analyse news in real time? AI does not know the events” (TV Journalist1 2024). Trends and Identification of Audience Preferences The use of AI-generated data to tailor content for audiences has a consid- erable impact on Albanian media. Among the 24 journalists and editors inter- viewed, 10 admitted to using data generated by AI to customise their content, while another 14 stated that they do not utilise AI for this purpose. It appears from the responses provided that AI is more commonly used by news portals than by television newsrooms for targeting audiences. Editors of portals acknow- ledge the use of AI for analysing audience behaviour and personalising content, whereas most television newsroom editors do not employ AI for this purpose. One television editor described AI as an invisible scanner that helps journal- ists identify topics of public interest by instantly analysing millions of posts on social networks and news sources, thereby accelerating the process of identifying important topics for immediate coverage. He articulated the role of AI in topic identification as follows: AI is that invisible scanner that moves faster than any reporter. With the ability to analyse millions of posts on social networks and news feeds in an instant, AI helps journalists discover breaking topics that require immediate coverage and that interest the public. (Tv Ed.1 2024) Two editors from online portals noted that AI is very useful for monitoring social networks and following public discussions, aiding newsrooms to discover trends and topics that attract more attention (Portal Ed.1 2024; Portal Ed.2 2024). AI can rank news based on relevance and popularity, facilitating the selection of 631 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 631 • let. 62, 3/2025 the most pertinent content for audiences. “ AI tools can follow posts and reactions on social platforms, discovering what is trending and how the public is perceiv- ing a certain event” (Portal Ed.1 2024). One interviewee noted: “By processing every interaction, every click, and every comment, it gives us a clear picture of the interests and preferences of the audience. AI is today an exceptional reader of the audience” (Tv Ed.1 2024). Another journalist uses AI to analyse audience demographics and interests, permitting the creation of more targeted content (Dig. Journalist1 2024). Some newsrooms leverage AI to test headlines and understand which topics resonate the most with audiences. “It is interesting to test the titles to see which ones per- form best, particularly on social networks and the television website” (Tv Ed.6 2024). One editor-in-chief of a television newsroom stated that AI has influenced the adaptation of content according to audience preferences. He asserted, “There is no longer room for random content. Each article is based on what readers are looking for, and AI specifically aids in this sophisticated adaptation” (Tv Ed.1 2024). Conversely, slightly more established applications and platforms (such as Google Analytics and Google Trends) remain the most commonly used tools for understanding audience behaviour and targeting content. One journalist from a portal explained, “AI helps collect and analyse data to segment audiences based on their behaviour, which aids in creating content tailored to specific groups” (Dig. Journalist3 2024). Yet, some newsrooms still do not extensively utilise AI for this function, relying more on traditional methods to follow news trends while, for others, AI provides general assistance. A journalist from an informational portal stated: “AI quickly processes the material, uses metaphors, and maintains harmony in the texts, but sometimes this material contains inaccuracies and can appear to be made up” (Dig. Journalist6 2024). Challenges and Expectations with Integrating AI into Editorial Processes From the perspective of media professionals, the use of AI in Albanian news- rooms brings several significant challenges. The lack of accuracy and reliability of the information generated by AI remains the central concern for many news- rooms. A considerable number of interviewees stressed that manual verification is still necessary to avoid errors. This highlights the prevailing mistrust among journalists and newsroom managers in new technologies, notably AI. One journ- alist observed, “The integration of artificial intelligence into editorial processes presents challenges for journalists, especially the older ones, related to a lack of technological knowledge” (Dig. Journalist1 2024). He stated that another chal- lenge is that AI does not possess as much information in the Albanian language as it does in other major languages, particularly English, which has added to the 632 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 632 TEORIJA IN PRAKSA scepticism among experienced journalists who have a specific memory of local events. Meanwhile, journalists recognised the need for training to enhance their knowledge and skills so as to facilitate the efficient use of AI by the media. Overall, the main challenges are related to accuracy, and the lack of context and adaptation to local realities. In terms of expectations, some journalists expressed the view that AI often fails to grasp the nuances of complex issues. DISCUSSIONS The use of AI by the Albanian media reflects different approaches among editors and journalists, as well as between television media and online portals. They regard AI as important in the research and verification stages of news pro- duction through various innovations. In answer to the first research question, it is evident that Albanian media currently employ AI for translation, transcrip- tion, text-to-speech conversions, summarising longer texts, gathering informa- tion, and generating ideas. Moreover, some advantages of using AI were identified, suggesting that the presence of non-human factors in the newsroom complements the human element. There is a category of editors and journalists who have embraced the impact of AI, while another group does not see it as a significant factor. Out of the 24 interviewees, 7 stated that AI has had a positive impact, improving content personalisation, trend analysis, and optimisation for different platforms. Conversely, 17 indicated that AI technology has not significantly affected editor- ial content since they continue to rely on traditional content-creation methods. Various studies indicate that journalists perceive automated journalism as a support mechanism that complements the role of human journalists. This view posits that automating routine processes in journalism permits journalists to focus more on engagements requiring unique human expertise and skills, such as original and in-depth reporting (Diakopoulos 2019). In many instances, journalists see AI as a pathway to integrating automated news writing within the news-making process (Thäsler-Kordonouri and Barling 2023). The expert - ise held by professional journalists is indispensable for ensuring that new tech- nologies are harmoniously integrated into existing news production structures (Milosavljević and Vobič 2019) and developed in line with the values of journal- ism, while maintaining its function of serving the public (Jones and Jones 2019). As concerns the role of AI in the quality and verification of information addressed in the second research question, Albanian journalists are divided in their opinions. Journalists and television editors perceive the human factor as irreplaceable because journalists provide a narrative context to information that AI cannot. The processing of large amounts of information by AI may also hold ethical implications. The emergence of AI-generated content in digital for- ums has sparked debates on legal and ethical issues, such as the risks of misin- formation and copyright protocols related to the data fed into generative models 633 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 633 • let. 62, 3/2025 (Thomson et al. 2024). Studies on the strategic use of misinformation in political discourse reveal that discussions about the problematic nature of information involving both political actors and journalists influence the determination of information legitimacy, identifying certain communication practices as harmful (Egelhofer et al. 2020). In general, Albanian journalists and editors maintain the position that the human factor is irreplaceable for verifying the information collected by AI, especially regarding new information and journalistic standards. Two editors of online portals opined that while AI is a useful tool for expediting and facilitat- ing processes, the verification and interpretation by journalists remain crucial to avoid misunderstandings and assure an accurate and reliable final product. Ana- lysis of media professionals’ attitudes suggests that cooperation between human expertise and technology can ensure that quality and accurate information is provided to the audience. To some extent, this position aligns with the findings of similar studies in other Western countries, which also stress that generative AI should not be per- mitted to operate without human oversight due to doubts regarding its accuracy, compliance with journalistic standards, and limitations in creativity and stylistic expression (Guenther 2025). AI can be utilised both ethically and unethically, but there is a collective need to strive for a new synthesis between our natural environments and digital technologies, focusing on how this new integration can support and develop a better society and a healthier biosphere (Floridi 2023). As for the third question, which pertains to the trend and identification of public preferences, Albanian media (both portals and television) are increasingly utilising AI to identify trends and public preferences. Analysis of the interviews indicates that the use of AI for this purpose has increased in nine newsrooms. Journalists and editors highlighted that AI aids with processing every click, comment and interaction to analyse audience interests and personalise content. Those who do not employ AI tend to rely more on traditional content-creation methods. This suggests that portals are more inclined to use AI for analysing and tailoring content according to audience interests. Namely, this is because the public does not constitute a homogeneous entity but is a diverse and fragmented audience that cannot be effectively reached through simplistic or uniform mes- saging (Kamin and Perger 2021). The mentioned orientation aligns with previous studies indicating that media focus on attracting larger audiences to maximise their product distribution (Blassnig and Esser 2021). A study conducted in Albania and Kosovo reveals that journalists and editors in those countries were inclined to be guided in their decision-making by audience data (Camaj et al. 2023). Another study emphas- ises that while television outlets in those countries continue to maintain an elite-centric approach to journalism, digital journalists, online newspapers, and portals utilising social media have adopted audience-centric attitudes to news (Saliu et al. 2024). 634 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 634 TEORIJA IN PRAKSA The other group consists of media professionals who continue to prioritise traditional approaches, maintaining editorial control and utilising AI primarily for data processing or initial information gathering. They acknowledged that AI has facilitated the collection and processing of information but asserted that it has not impacted editorial content. These professionals advocate an elite-centric approach to journalism focused on important societal developments rather than public entertainment (Saliu et al. 2024). Some findings – such as the use of AI for transcription, translation, and to improve time-efficiency – may seem predictable in well-resourced media sys- tems. However, their recurrence in this study should be understood within the structural limitations of Albanian journalism. These practices reflect more than operational convenience; they show how journalists adapt to automation in the absence of institutional support. From the perspective of technological affordances (Norman 1999), these tools are not merely available but selectively activated in response to professional constraints. The tendency to deploy AI in low-risk tasks while retaining human control over editorial decisions con- sequently aligns with the logic of hybrid newsrooms (Newman 2022; Hendrickx and Picone 2024) where journalistic authority continues to hold precedence over algorithmic output. This dynamic suggests a cautious, context-specific integra - tion of AI rather than any full-scale transformation of news production. While AI may improve the working conditions of journalists, it should not be permit- ted to operate without human oversight (Guenther 2025). Importantly, the trends observed in Albania are not isolated. Similar pat- terns have been detected in other Central and Eastern European contexts. For example, a study conducted in countries such as the Czech Republic, Hungary, Poland and Slovakia reveals that editorial offices across these nations share com- parable concerns – seeking to balance technological innovation with ethical journalism practices (Thomson Foundation and Media and Journalism Research Center 2024). These regional findings also reflect broader global research, which reveals persistent disparities in AI adoption, particularly in transitional demo- cracies and smaller media markets (Beckett and Yaseen 2023). CONCLUSIONS Based on data collected from 24 interviews with editors and journalists from Albanian television and portal newsrooms, we conclude that the use of AI in Albanian newsrooms is generally still in its infancy. The adoption of AI in Albanian media is on the rise, bringing several advantages, including time optimisation and improved news quality. Journalists utilise AI for automatically writing simple news stories, translating materials, and transcribing interviews, allowing them more time for in-depth analysis. Still, the reliability of AI-gener- ated information remains a concern. Most journalists view the human role as irreplaceable for verifying and interpreting information. 635 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 635 • let. 62, 3/2025 The use of AI for analysing audience behaviour is prevalent in online portals, whereas TV newsrooms are more cautious. Continuous improvement in tech- nological knowledge and training for journalists is required to address the chal- lenges of integrating AI into editorial processes. Despite the many advantages, scepticism remains as to the accuracy and effi- ciency of AI, particularly for current information, as these technologies often rely on outdated data. Journalists and editors assert that ensuring high journal- istic quality calls for collaboration between AI and media professionals. The integration of AI into journalism requires a careful approach that assures not just efficiency but also the reliability and accuracy of information, thereby maintaining the journalist’s central role in reporting and verifying news. Concerns arise regarding the adoption of AI in journalism, particularly regarding algorithmic biases, maintaining editorial independence, upholding journalistic values, ensuring transparency in news production processes, chal- lenges in fact-checking, data ethics, and the essential role of human judgment in decision-making. In summary, although AI aids in supporting processes and daily journalistic routines, its impact on editorial content remains moderate. Limitations and Future Research This study has certain limitations, mostly due to its qualitative design. The sample is not statistically representative, and the findings are context-specific rather than broadly generalisable. However, the inclusion of a diverse group of journalists and editors from both television and digital platforms strengthens the study’s ability to provide insights into how AI is conceptualised and imple- mented in transitional media environments. Future research could explore the long-term impact of AI adoption on journ- alistic values and routines, especially in terms of how it influences editorial judgment, public trust, and working conditions. Comparative studies across countries with transitional media systems may offer valuable perspectives on structural differences in the integration of AI into journalism. Moreover, audi- ence-centred research is essential for understanding public perceptions of, and reactions to, the content produced or assisted by AI technologies. BIBLIOGRAPHY Ali, Waleed, and Mohamed Hassoun. 2019. “Artificial Intelligence and Automated Journalism: Contemporary Challenges and New Opportunities.” International Journal of Media, Journ- alism and Mass Communications 5 (1): 40–9. https://doi.org/10.20431/2454-9479.0501004. Allcott, Hunt, and Matthew Gentzkow. 2017. “Social Media and Fake News in the 2016 Election.” Journal of Economic Perspectives 31 (2): 211–36. https://doi.org/10.1257/jep.31.2.211. Al-Zoubi, Omar, Normahfuzah Ahmad, and Norsiah, Abdul Hamid. 2024. “Artificial Intelligence in Newsrooms: Ethical Challenges Facing Journalists.” Studies in Media and Communica- tion 12 (1): 401–09. https://doi.org/10.11114/smc.v12i1.6587. 636 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 636 TEORIJA IN PRAKSA Barclay, Donald A. 2018. Fake News, Propaganda, and Plain Old Lies: How to Find Trustworthy Information in the Digital Age. London: Rowman and Littlefield. Barrolleta, Leonardo Alberto La-Rosa, and Teresa Sandoval-Martín. 2024. “Artificial Intelli- gence versus Journalists: The Quality of Automated News and Bias by Authorship Using a Turing Test.” Anàlisi: Quaderns de Comunicació i Cultura 70: 15–36. https://doi.org/10.5565/ rev/analisi.3681. Beckett, Charlie, and Mira Yaseen. 2023. “Generating Change: A Global Survey of What News Organisations Are Doing with AI.” JournalismAI. https://www.journalismai.info/re- search/2023-generating-change. Biswal, Santosh Kumar, and Nikhir Kumar Gouda. 2020. “Artificial Intelligence in Journalism: A Boon or Bane?” In: Optimization in Machine Learning and Applications, Ashok J. Kulkarni and Subhendu Chattaraj Satapathy (eds.), 155–67. Springer. https://doi.org/10.1007/978- 981-15-0994-0_10. Blassnig, Stefan, and Frank Esser. 2021. “The ‘Audience Logic’ in Digital Journalism: An Explor- ation of Shifting News Logics Across Media Types and Time”. Journalism Studies 23 (1): 48–69. https://doi.org/10.1080/1461670X.2021.2000339. Braun, Virginia, and Victoria Clarke. 2006. “Using Thematic Analysis in Psychology.” Qualitative Research in Psychology 3 (2): 77–101. https://doi.org/10.1191/1478088706qp063oa. Camaj, Lindita, Erlis Çela, and Gjylie Rexha. 2023. “The Audience Logic in Election News Re- porting on Facebook: What Drives Audience Engagement in Transitional Democracies of Albania and Kosovo?” Journal of Information Technology and Politics 21 (1): 38–53. https:// doi.org/10.1080/19331681.2023.2227847. Cools, Henk, and Nicholas Diakopoulos. 2024. “Uses of Generative AI in the Newsroom: Map- ping Journalists’ Perceptions of Perils and Possibilities.” Journalism Practice, 1–19. https:// doi.org/10.1080/17512786.2024.2394558. Cools, Henk, and Nicholas Diakopoulos. 2024. “Uses of Generative AI in the Newsroom: Map- ping Journalists’ Perceptions of Perils and Possibilities.” Digital Journalism, 1–19. https://doi. org/10.1080/17512786.2024.2394558. Cools, Henk, Baldwin Van Gorp, and Michael Opgenhaffen. 2022. “Where Exactly Between Utopia and Dystopia? A Framing Analysis of AI and Automation in US Newspapers.” Journ- alism, 1–19. https://doi.org/10.1177/14648849221122647. Dalen, Andre Van. 2024. “Revisiting the Algorithms Behind the Headlines: How Journalists Respond to Professional Competition of Generative AI.” Journalism Practice, 1–18. https:// doi.org/10.1080/17512786.2024.2389209. Denzin, Norman K., and Yvonna S. Lincoln. 2017. The Sage Handbook of Qualitative Research (5th ed.). Thousand Oaks, CA: Sage. Deuze, Mark, and Charlie Beckett. 2022. “Imagination, Algorithms and News: Developing AI Literacy for Journalism.” Digital Journalism 10 (10): 1913–18. https://doi.org/10.1080/21670 811.2022.2119152. Deuze, Mark, and Tamara Witschge. 2019. Beyond Journalism. Cambridge: Polity Press. Diakopoulos, Nicholas, and Michael Koliska. 2016. “Algorithmic Transparency in the News M e d i a .” Digital Journalism, 1–21. https://doi.org/10.1080/21670811.2016.1208053. Diakopoulos, Nicholas. 2019. Automating the News: How Algorithms are Rewriting the Media. Cambridge: Harvard University Press. Dodds, Tomas, Claes de Vreese, Natalie Helberger, Valeria Resendez, and Theresa Seipp. 2023. “Popularity-driven Metrics: Audience Analytics and Shifting Opinion Power to Digital Plat- 637 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 637 • let. 62, 3/2025 forms.” Journalism Studies 24 (3): 403–21. https://doi.org/10.1080/1461670X.2023.2167104. Egelhofer, Johannes Laura, Loes Aaldering, Jakob-Moritz Eberl, and Sebastian Galyga. 2020. “From Novelty to Normalization? How Journalists Use the Term ‘Fake News’ in Their R e p o r ti n g . ” Studies in Media and Communication 21 (10): 1323–43. https://doi.org/10.1080 /1461670X.2020.1745667. European Commission. 2021. Ethics and Data Protection. Brussels: European Commission. h t tp s : / / e c . e u r o p a . e u / i n f o / s i t e s / d e f au l t / f i l e s / 5 . _ h 2 0 2 0_ e th i c s _ a n d _ d a t a _ p r o t e c ti o n _ e n . p d f . European Commission. 2023. Key Findings of the 2023 Report on Albania. Brussels: European Commission. https://ec.europa.eu/commission/presscorner/detail/en/qanda_23_5612. Floridi, Luciano. 2023. The Ethics of Artificial Intelligence: Principles, Challenges, and Oppor- tunities. Oxford: Oxford University Press. Graefe, Andreas. 2016. “Guide to Automated Journalism”. Tow Center for Digital Journalism report. Accessed 18 September 2024. https://data.journalism.columbia.edu/sites/default/ files/content/guide-to-automated-journalism.pdf. Guenther, Lars, Jessica Kunert, and Bernhard Goodwin. 2025. “My New Colleague, ChatGPT? How German Science Journalists Perceive and Use (Generative) Artificial Intelligence.” Journalism Practice 19 (10): 2177–88. https://doi.org/10.1080/17512786.2025.2502794. Hendrickx, Jonathan, and Ike Picone. 2024. “Operationalising Hybrid Newsroom Ethnography: Observing Amidst a Pandemic.” Journalism Practice 18 (4): 840–57. https://doi.org/10.108 0/17512786.2022.2062429. Hölsgens, Sander, Saskia de Wildt, and Tamara Witschge. 2020. “Towards an Experientialist Understanding of Journalism: Exploring Arts-based Research for Journalism Studies.” Journalism Studies 21 (7): 928–46. https://doi.org/10.1080/1461670X.2019.1703121. Houston, Brant. 2023. Changing Models for Journalism: Reinventing the Newsroom. New York and London: Routledge. Illia, Laura, Elanor Colleoni, and Stelios Zyglidopoulos. 2023. “Ethical Implications of Text Generation in the Age of Artificial Intelligence.” Business Ethics: A European Review 32 (1): 201–10. https://doi.org/10.1111/beer.12479. Jamil, Sadia. 2020. “Artificial Intelligence and Journalistic Practice: The Crossroads of Obstacles and Opportunities for the Pakistani Journalists.” Journalism Practice 15 (10): 1400–22. ht t p s : / / d o i . o rg / 1 0 . 1 0 8 0 / 17 5 1 2 7 8 6 . 2 02 0 . 17 8 8 4 1 2 . Jones, Bronwyn, and Rhianne Jones. 2019. “Public Service Chatbots: Automating Conversation with BBC News.” Digital Journalism 7 (8): 1032–53. https://doi.org/10.1080/21670811.2019. 1609371. Kamin, Tanja, and Nina Perger. 2021. “Napake v krizni komunikaciji med pandemijo COVID-19 v Sloveniji: Kaj nas lahko nauči odziv publike.” Teorija in praksa 58: 517–35. https://doi. org/10.51936/tip.58.specialissue.517-535. Karim, Schapals Aljosha, and Colin Porlezza. 2020. “Assistance or Resistance? Evaluating the Intersection of Automated Journalism and Journalistic Role Conceptions.” Media and Com- munication, 8 (3): 16–26. https://doi.org/10.17645/mac.v8i3.3054 Kovach, Bill, and Tom Rosenstiel. 2021. The Elements of Journalism: What Newspeople Should Know and the Public Should Expect. New York: Crown. Lamot, Kenza, and Steve Paulussen. 2020. “Six Uses of Analytics: Digital Editors’ Perceptions of Audience Analytics in the Newsroom.” Journalism Practice 14 (3): 358–73. https://doi.or g/10.1080/17512786.2019.1617043. Lee, Eun-Ju, and Edson C. Tandoc Jr. 2017. “When News Meets the Audience: How Audience 638 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 638 TEORIJA IN PRAKSA Feedback Online Affects News Production and Consumption.” Human Communication R e s e a rc h 43 (4): 436–49. https://doi.org/10.1111/hcre.12123. Lopez, Marisela Gutierrez, Colin Porlezza, Glenda Cooper, Stephann Makri, Andrew MacFarlane, and Sondess Missaoui. 2023. “A Question of Design: Strategies for Embedding AI-Driven Tools into Journalistic Work Routines.” Digital Journalism 11 (3): 484–503. https://doi.org/1 0.1080/21670811.2022.2043759. Media Ownership Monitor Albania. 2023. Key Findings. RSF and BIRN Albania. https://albania. mom-gmr.org/en/. Milosavljević, Marko, and Igor Vobič. 2019. “Human Still in the Loop: Editors Reconsider the Ideals of Professional Journalism Through Automation.” Digital Journalism 7 (8): 1098–1116. https://doi.org/10.1080/21670811.2019.1601576. Munoriyarwa, Allen, Sarah Chiumbu, and Gilbert Motsaathebe. 2023. “Artificial Intelligence Practices in Everyday News Production: The Case of South Africa’s Mainstream News- rooms.” Journalism Practice 17 (7): 1374–92. https://doi.org/10.1080/17512786.2021.19849 76. Nelson, Jacob L., and Edson C. Tandoc Jr. 2018. “Doing ‘Well’ or Doing ‘Good’: What Audience Analytics Reveal About Journalism’s Competing Goals.” Journalism Studies 20 (13): 1960– 76. https://doi.org/10.1080/1461670X.2018.1547122. Newman, Nic, Richard Fletcher, Anne Schulz, Simge Andı, Claire T. Robertson, and Rasmus Kleis Nielsen. 2021. Reuters Institute Digital News Report 2021. Oxford: Reuters Institute for the Study of Journalism. Newman, Nic, Richard Fletcher, Craig T. Robertson, Kirsten Eddy, and Rasmus Kleis Nielsen. 2022. Reuters Institute Digital News Report 2022. Oxford: Reuters Institute for the Study of Journalism. Newman, Nic. 2022. Journalism, Media, and Technology Trends and Predictions 2022. Ox- ford: Reuters Institute for the Study of Journalism. https://reutersinstitute.politics.ox.ac.uk/ sites/default/files/2022-01/Newman%20-%20Trends%20and%20Predictions%202022%20 FINAL.pdf. Norman, Donald A. 1999. “Affordance, Conventions, and Design.” Interactions 6 (3): 38–43. https://doi.org/10.1145/301153.301168. Pajo, Bora. 2018. Introduction to Research Methods: A Hands-On Approach. Los Angeles: Sage Publications Inc. Pavlik, John V. 2000. “The Impact of Technology on Journalism.” Journalism Studies, 1 (2): 229–37. Pavlik, John V. 2023. “Collaborating with ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education.” Journalism and Mass Commu- nication Educator 78 (1): 84–93. https://doi.org/10.1177/10776958221149577. Pavlin, Samo, and Ivan Svetlik. 2024. “ChatGPT, napiši nama članek o možnostih uporabe umetne inteligence v visokošolskem izobraževanju!” Teorija in praksa 61 (4): 847–64. https: //doi.org/10.51936/tip.61.4.847. Saliu, Hasan, and Venera Llunji. 2022. “Cultural Diplomacy of Kosovo after the Declaration of Independence.” Information and Media 93: 62–76. https://doi.org/10.15388/Im.2022.93.61. Saliu, Hasan, Ardita Reçi, and Gazmend Abrashi. 2023. “Revisiting Hallin and Mancini’s Me- dia Model: Albania and Kosovo.” Balkan Social Science Review 21: 235–55. https://doi. org/10.46763/BSSR2321235s. Saliu, Hasan, Ramadan Çipuri, and Xhevahire Izmaku. 2024. “Gutenberg’s Death in the Balkans: 639 • let. 62, 3/2025 • AI in the Newsroom: Changes and Challenges in Albania 639 • let. 62, 3/2025 News Values in Kosovo and Albania.” Cogent Arts and Humanities 11 (1): 2303197. https:// doi.org/10.1080/23311983.2024.2303197. Saliu, Hasan. 2023. “Narratives of Public Diplomacy in the post-Truth Era: The Decline of Soft Power.” Communication and Society 36 (2): 209–24. https://doi.org/10.15581/003.36.2.209- 224. Saliu, Hasan. 2024. “Navigating Media Literacy in the AI Era: Analyzing Gaps in Two Classic Media Literacy Books.” Journal of Applied Learning and Teaching 7 (2): 98–109. https:// doi.org/10.37074/jalt.2024.7.2.25. Simon, Felix M. 2024. “Escape Me If You Can: How AI Reshapes News Organisations’ Depend- ency on Platform Companies.” Digital Journalism, 12 (2): 149–70. https://doi.org/10.1080/2 1670811.2023.2287464. Srisaracam, Sakulsri. 2023. “Convergent Journalism. Cross-Media Content Strategies to Improve the Quality of Thai News Reporting.” The Routledge Companion to News and Journalism, Stuart Allan (ed.), 130–38. London and New York: Routledge. Tandoc Jr., Edson C., Lim Jia Yao, and Shangyuan Wu. 2020. “Man vs. Machine? The Impact of Algorithm Authorship on News Credibility.” Digital Journalism 8 (4): 548–62. https://doi.or g/10.1080/21670811.2020.1762102. Tandoc, Edson C. 2014. “Journalism is Twerking? How Web Analytics is Changing the Pro- cess of Gatekeeping.” New Media and Society 16 (4): 559–75. https://doi.org/10.1177/ 1461444814530541. Tandoc, Edson C., and Julian Maitra. 2017. “News Organizations’ Use of Native Videos on Face- book: Tweaking the Journalistic Field One Algorithm Change at a Time.” New Media and Society 20 (5): 1679–696. https://doi.org/10.1177/1461444817702398. Tariq, Muhammad, Muhammad Jawed Aslam, Abdul Shakoor, and Saba Ilyas. 2024. “Artificial Intelligence and the Reshaping of Journalism.” Qtlantic Journal of Social Sciences 5 (1): 44–53. Thäsler-Kordonouri, Sina, and Kurt Barling. 2023. “Automated Journalism in UK Local News- rooms: Attitudes, Integration, Impact.” Journalism Practice 19 (1): 58–75. https://doi.org/10. 1080/17512786.2023.2184413. Thomson Foundation, and Media and Journalism Research Center. 2024. How Artificial Intel- ligence Is Changing Media and Journalism in Central Europe: A Study Mapping the Use of AI by Newsrooms in the Czech Republic, Hungary, Poland and Slovakia. https://www. thomsonfoundation.org/media/269005/tf_ai_in_v4_newsrooms.pdf. Thomson, T. J., Ryan J. Thomas, and Phoebe Matich. 2024. “Generative Visual AI in News Or- ganizations: Challenges, Opportunities, Perceptions, and Policies.” Digital Journalism, 1–23. https://doi.org/10.1080/21670811.2024.2331769. Verčič, Dejan. 2022. “Na začetku in koncu demokracije je (bila) beseda: O politologiji in komunikologiji.” Teorija in praksa 59 (4): 378–92. Wang, Quan. 2017. “Dimensional Field Theory: The Adoption of Audience Metrics in the Journ- alistic Field and Cross-Field Influences.” Digital Journalism 6 (4): 472–91. https://doi.org/10 .1080/21670811.2017.1397526. Zhang, Wei, José Manuel Pérez Tornero, and Qiusheng Tiani. 2023. “Dissecting Automated News Production from a Transdisciplinary Perspective: Methodology, Linguistic Applica- tion, and Narrative Genres.” Sage Open 13 (4). https://doi.org/10.1177/21582440231206372. 640 TEORIJA IN PRAKSA • Ramadan ÇIPURI, Erlis ÇELA, Hasan SALIU 640 TEORIJA IN PRAKSA UMETNA INTELIGENCA V NOVINARSKIH REDAKCIJAH: SPREMEMBE IN IZZIVI V ALBANIJI Povzetek. Članek raziskuje vpliv umetne inteligence (UI) na novinarske re- dakcije v Albaniji s pomočjo kvalitativnega raziskovalnega pristopa. Na podlagi 24 polstrukturiranih, poglobljenih intervjujev z novinarji in uredniki so preučene izkušnje in dojemanja medijskih strokovnjakov glede vključevanja UI v njihovo delo. Rezultati kažejo, da se UI uporablja za zbiranje informacij, prevajanje, pre- verjanje slik in analizo trendov v novicah. Med prednostmi uporabe UI sta iz- postavljena časovna učinkovitost in izboljšanje kakovosti novic, kljub temu pa ostajajo izzivi, kot so skrb za točnost informacij in dvom v tehnologijo. Novinarji poudarjajo potrebo po ročnem preverjanju ter nenadomestljivo vlogo človeškega dejavnika pri zagotavljanju varnosti informacij. Uporaba podatkov UI za prepo- znavanje interesov občinstva prav tako kaže na večjo sposobnost UI pri ustvarja- nju vsebin, ki pritegnejo bralce. V študiji je poudarjeno, da je sodelovanje med UI in novinarji ključno za zagotavljanje kakovosti in natančnosti v novinarstvu. Ključni pojmi: umetna inteligenca, UI, novinarstvo, mediji, družbena omrež- ja, novice, redakcija.