For over two centuries, the Armenian news media has evolved with the times to accomplish one overarching goal: to keep Armenians connected to their history, identity, heritage, and most of all, each other. Thanks to technological advancements, the field is more diverse and far-reaching than ever—from local newspapers, regional broadcasts, social media platforms and independent podcasts to webcasts and satellite feeds.
Now, in the age of artificial intelligence, news and information outlets are caught in a struggle between content that is created, simulated, or manipulated by synthetic or digital means and the natural, objective events that occur independently of our conscious creation. Yet these new super tools of production can easily shapeshift into the news itself for audiences to discover, consume, and trust, often at their own peril.
In South America, Armenian communities have deep roots dating back to the waves of Genocide survivors who settled in Argentina, Uruguay, and Brazil in the early twentieth century.
Founded in 1975, Sardarabad is one of the most influential Armenian diaspora media outlets in Latin America. Based in Argentina, it has evolved from a print publication into a fully digital platform, publishing content in Spanish, Armenian, English, and Portuguese. The publication carries both a historical weight and a living mission: to preserve and transmit Armenian identity, memory, and culture to communities scattered across the world. Editor-in-Chief Sergio Nahabetian sees AI as the natural next step in that long journey of adaptation, arguing that “refusing to use it would be equivalent to resisting, decades ago, the arrival of the internet or mobile phones.”
Yet Nahabetian is equally clear about where the limits lie. He warns that “there is also the risk of misinformation. AI can reproduce errors, incomplete data, or biased content if there is no serious editorial work behind it to verify sources and provide proper judgment. In matters related to historical memory, identity, and diaspora, this becomes especially delicate.” For a publication so deeply tied to collective memory, Nahabetian believes technology is only as responsible as the editorial leadership guiding it and that “editorial judgment, cultural experience, and the knowledge of those who are part of the publication remain irreplaceable.”
Audición Gomidas in Uruguay is a good example. Founded in 1935, it is the oldest Armenian radio broadcast in the Diaspora. The station continues today under a new generation of young Uruguayan Armenians. Its programming blends music, interviews, commemorations, and community news in what General Coordinator and Host Romina Kuyunjian describes as “a dynamic and contemporary style.”
The station’s use of AI has developed gradually. Today, it is primarily used for such functions as editing content, creating scheduling spreadsheets for hosts, and organizing news and information from multiple sources. As Kuyunjian explains, “AI saves time and allows us to focus more on editorial judgment and message-building.”
Nonetheless, she is clear about its limitations, noting, “AI can provide information, but it cannot grasp who we are or the significance our radio station holds for our community.” The Armenian story is extensive, complex, and deeply symbolic, shaped by lived experience and cultural nuance that cannot be easily captured through existing datasets.
Kuyunjian points to concerns about the loss of authenticity, overreliance on automation, and the spread of unverified content. The priority, she emphasizes, is maintaining balance: using technology as a tool, while ensuring that human judgment, cultural understanding, and community responsibility remain at the core of the station’s work.
A Two-Sided Story
In the Middle East, another traditional news publication that experiences a similar struggles is Aztag (Impulse/Stimulus). A daily newspaper based in Bourj Hammoud, Lebanon, it has been publishing since 1927 against every possible odd. It circulates six days a week, and is read not only throughout Lebanon but by Diasporans worldwide. Under the leadership of Editor-in-Chief Shahan Kandaharian since 2002, Aztag has weathered Lebanon’s Civil War, the COVID-19 pandemic, the Beirut Blast of 2020 along with one of the worst economic collapses in the country, not to mention the recent foreign invasions into Lebanon.
Kandaharian speaks with conviction about why hard news still matters and traditional print media in particular. “In a place like Lebanon, many older generation Armenians wake up and reach for Aztag, Zartonk or Ararad as they did for years before the digital revolution. The haptic experience of holding the paper, turning its pages, reading in their mother tongue and sharing feedback with friends and neighbors is a morning ritual that carries a weight that no digital notification can replicate.”
Yet on the question of AI, he concedes real benefits, especially in archiving: “Aztag has nearly a century of issues in its archives, and AI-powered tools could transform that into a searchable, living repository of Armenian history and community life. Our editorial staff is encouraged to use AI for tasks such as generating summaries and suggesting headlines and story angles.”
However, he is emphatic that AI be used as a tool of production not as the last word, explaining that efficiency in process can compromise the integrity of the content that is essential to helping audiences make informed decisions about their lives and their communities.
With Western Armenian already fighting for survival, heavy reliance on AI tools, without rigorous editorial oversight, could quietly erode the very language Armenian media seeks to protect.
He is also concerned about how AI writing and translation tools are trained overwhelmingly on dominant languages such as English, Spanish, and French. Western Armenian, which UNESCO classifies as critically endangered, is dramatically underrepresented. “The result is AI-generated Armenian that is often grammatically imprecise and culturally tone-deaf,” he notes. “With Western Armenian already fighting for survival, heavy reliance on these tools, without rigorous editorial oversight, could quietly erode the very language Armenian media seeks to protect.”
In Armenia, broadcast TV has evolved significantly since independence in 1991, when state broadcasting gave way to a more pluralistic—if still politically complex—media landscape. Public Television of Armenia remains the country’s main public broadcaster, while channels such as Armenia TV, Shant TV, Yerkir Media, and ATV serve large domestic and Diaspora audiences. Internationally, digital distribution has expanded their reach considerably, with many channels now streaming live and making their archives available to Diaspora viewers worldwide.
Tsaghkanush Mirzoyan, a journalist at Public Television of Armenia, captures the medium’s cautious relationship with AI. She uses it for preliminary research, prompting GPT to surface key articles on a topic, but insists on returning to the source material herself: “I still research using the original material and read the actual articles, but it saves me a great deal of time in the search phase.” Editing and scripting, however, remain entirely traditional. Her team still spends forty minutes to an hour on a three-minute report, because “AI language is very predictable and tends to lengthen sentences, while in TV journalism every word counts. We need to convey as much information as possible within a limited word count.” For Armenian specifically, the gap is even more pronounced: AI, she explains, still applies an English-language mentality to Armenian, making machine-generated text immediately detectable to any native reader.
The broader promise of AI for Armenian television is nonetheless real: automated subtitling can make content available in multiple languages within hours, real-time translation could reconnect Diaspora viewers who have lost fluency in Armenian, and decades of archives could become searchable through AI-powered indexing. Yet the risks mirror those in print and radio—overreliance on machine translation that lacks cultural nuance, younger generations consuming Armenian content exclusively through subtitles rather than engaging with the language itself, and the gradual erosion of fluency as AI removes the incentive to preserve it.
Podcast Boom
Armenian podcasting has flourished across a remarkable range of voices and topics. Podmootiun (a pun on the Armenian word for history, badmutiun) brings centuries of Armenian heritage to life through conversations of Father Tadeos Barseghyan and Peter Hajinian. Jazve Talks has carved out a feminist space, a self-described “amot-free zone” for honest conversations with Armenian artists, writers, and creatives. EVN Report has extended its analytical journalism into audio as well. Institutions have joined the conversation too. The USC Dornsife Institute of Armenian Studies produces podcast series exploring language, Diaspora identity, and policy from an academic perspective.
What is perhaps most striking about the Armenian podcasting landscape is how democratizing it has been. The above-mentioned traditional Armenian media requires institutional backing, infrastructure, funding and significant resources. A podcast requires a microphone, an internet connection, and something worth saying.
Seda Arzumanyan, founder of the U.S.-based Beyond Borders Diaspora podcast, is candid about what made her idea a reality: AI tools—most notably Riverside, a professional podcast recording platform are central to her workflow.
“Without AI tools like Riverside, this podcast never would have launched. The technology handles editing, transcription, and video clipping that would otherwise require a full production team. It’s democratized podcasting for the Armenian diaspora, allowing individual voices to reach global audiences with professional quality,” shares Arzumanyan. For her, producing a podcast polished and credible enough to represent the Armenian diaspora would have otherwise required substantial funding. Riverside removed those barriers.
Not every podcast, however, is built for the same kind of AI integration. The Focus Series, produced by the Media Lab at the American University of Armenia, is an English-language podcast that brings together academics, professionals, alumni, and students to discuss topics ranging from politics and public health to media and culture—all in service of the Lab’s broader mission to foster democratic culture and journalism in Armenia. Its structure, where two experts drive an unscripted conversation without a fixed host, shapes how AI fits into the workflow. As AUA Chief Communications Officer Narek Ghazaryan explains, “Since the structure of the podcast is unique and it does not have a fixed host, we rarely turn to AI in designing the script or the plot of the episode. Conversation is usually between two experts of the field where one is the host and the other is the guest who both develop the discussion in a natural manner.” That said, AI is not absent entirely: “In a podcast’s post-production phase, we have incorporated new AI driven techniques that make various editing processes more efficient.” The Focus Series and Beyond Borders Diaspora together illustrate a broader truth about AI in Armenian podcasting: its value depends entirely on the nature of the project, and the most thoughtful producers are the ones who know exactly where to let it in.
The challenge for Armenian media is to use AI without surrendering to its limitations; to embrace efficiency without sacrificing authenticity; to reach new audiences without diluting nuance and losing sight of why those audiences sought out Armenian media in the first place.
Misinformation Machine
The World Economic Forum has identified AI-driven misinformation as one of the most severe global risks in the years ahead. While AI has expanded content creation, it is also accelerating the spread of false news. Misinformation is inaccurate information spread without malicious intent. The source presumes something to be true, unaware it is based on rumor, an incorrect statistic or misinterpretation of a news story. With AI systems trained on vast amounts of unreliable online data, inaccuracies are easily reproduced and amplified. In contrast, disinformation is fake news deliberately created and spread with the intent to mislead, harm or manipulate. This includes malicious content like propaganda, hoaxes and deepfakes.
Both pose challenges that are particularly acute for smaller media ecosystems like Armenian news. Research shows that AI detection and verification tools are less effective in non-dominant languages, creating an imbalance where misinformation spreads more easily than it can be fact-checked. At the same time, many newsrooms lack formal training in responsible AI use, increasing the risk of unintentional errors.
Yet the impact is not purely linear. Studies suggest AI-generated misinformation tends to reinforce existing beliefs rather than radically shift them. In this environment, credibility becomes more valuable: as false content proliferates, audiences may place greater trust in established, reliable sources—raising the stakes for responsible journalism.
Proceed with Caution
The story of Armenian news media in the age of AI is not one of simple technological adoption or resistance. It is a story of communities trying to preserve language, culture, and connection while navigating tools that promise efficiency but threaten authenticity; that offer reach but risk dilution; that enable creation but also enable deception.
Kandaharian sums it up, saying, “The challenge for Armenian media is to use AI without surrendering to its limitations; to embrace efficiency without sacrificing authenticity; to reach new audiences without diluting nuance and losing sight of why those audiences sought out Armenian media in the first place.”
Whether the community’s media outlets succeed in striking that balance will help determine whether Armenian language and culture remain vibrant or slowly fade into digital noise.
