In recent years, traditional media houses across the globe have resorted to layoffs as a last-ditch effort to stay afloat; if not for the long term, then at least for a few more years. This wave of downsizing began during the Covid-19 crisis and has yet to subside.
Media organizations are now restructuring into smaller, smarter and more agile newsrooms to cut costs. They are grappling with a severe financial crisis as conventional revenue streams dry up and new ones are slow to emerge. Nepal is no exception to this trend. To reduce expenses, many media houses are working to merge operations across print, radio, television and digital platforms into unified newsrooms. The only seemingly viable, though not well-thought-out option has been to scale down operations to match dwindling revenues.
In this context, a wide range of Artificial Intelligence (AI) tools could prove to be a boon for the fragile media landscape, potentially helping to fill gaps left by staff reductions. However, before embracing AI more broadly, it is crucial for media houses to formulate clear policies to ensure its ethical, transparent and effective use.
While some media houses have already started using AI tools, their applications remain minimal and largely unregulated. It is high time media houses moved decisively, from the Gutenberg-era newsroom to an AI-equipped, high-tech newsroom. A key first step in this transition is to provide training for journalists and collaborate with technology companies to develop customized newsroom tools. While the adoption of AI is not without costs, it can be a cost-effective alternative in the long run, gradually replacing outdated editorial structures.
At present, AI use in Nepal’s newsrooms is limited to individual journalists. Many AI-generated, translated or edited texts are published without any editorial supervision. While no comprehensive study has been conducted to assess the use of AI in Nepali newsrooms so far, a recent survey by Rajiv Timalsina, a student of Kantipur City College, provides some insights. According to the survey, 38 percent of journalists use AI tools for transcription tasks such as documenting interviews. Around 22 percent use them for fact-checking through platforms like Google Fact Check or image verification engines, while only 18 percent use audience analytics tools to understand reader behavior.
The next step after introducing AI tools is to train employees to use them effectively, particularly to improve the quality of writing and editing. Currently, there is a lack of trained human resources in the newsroom, and local journalists are still in the early stages of AI adoption. While some non-governmental organizations have begun offering training, there has been little to no institutional collaborations.
Media houses must establish dedicated AI departments and AI editors to provide proper insight and guidance. Without this, the unchecked use of AI could lead to serious problems. If possible, Nepali media should also seek collaboration with international media organizations to learn from their experiences, though even global media outlets are still experimenting with AI integration.
In 2024, The New York Times publicly released a document outlining its approach to AI in the newsroom. The US media company said it does not use AI to write news or articles. It said it uses AI in three main ways: as a tool in the service of its journalistic mission, under human guidance and review and transparent use. Compared to other international media outlets, it has adopted a more cautious stance on AI use, maintaining that human creativity remains central to content creation.
In 2023, The Financial Times appointed Madhumita Murgia as its first AI editor. The following year, The New York Times rolled out its first generative AI features for subscribers. The same year, The Washington Post launched “Ask the Post AI”, which it described as a generative AI tool leveraging the publication’s deeply-sourced, fact-based journalism to deliver summary answers and curated results directly to users.
In neighboring India, The Hindustan Times joined the AI race in 2024, establishing a 15-member team to work on GenAI-based initiatives such as news bots, personalization, audience engagement, monetization and subscription strategies.
In Nepal, while journalists have begun using publicly-available AI tools, institutional adoption remains limited. However, some media outlets have started integrating AI technologies in various ways, from digital news readers to audio transcription, translation, image and text generation, and even news writing.
Providing summaries alongside news articles, with editorial endorsement, is a growing global trend. Onlinekhabar is among the Nepali media following this trend. Annapurna Post has also taken help of AI for its digital reader tool. However, some news outlets, which lack strong editorial oversight, are publishing AI-generated summaries that are flawed or misleading.
With the use of AI at the individual level growing, media houses must ensure that AI is used responsibly and ethically. This is necessary both to maintain editorial integrity and to earn people’s trust. With strategic investment and collaboration with tech companies, AI could unlock new opportunities for Nepali media houses.
AI tools can help summarize news stories, but editors must have the final say. The New York Times’ own experience shows that AI-generated summaries often fail to fully capture the nuances of original articles. Many believe that human-written summaries are still superior. Despite this, the US media company’s AI team has been refining its tools, acknowledging that while AI is not perfect, it can still help free up editorial staff for other important tasks.
With the right human guidance, generative AI can also be used to create visuals to accompany new stories. Some Nepali media houses have started using AI-generated images, but these are only accurate when journalists provide detailed guidance. Otherwise, there is a risk of misrepresentation and factual inaccuracies.
Resource constraints have long prevented Nepal media from producing investigative, analytical and in-depth news stories. AI could help bridge this gap. It can assist in scanning documents, analyzing data and identifying leads for investigative reports. These are the tasks that journalists often struggle to manage under tight deadlines. AI can also support wider and more efficient coverage by translating news into multiple languages to reach broader audiences. Some outlets in Nepal have already begun experimenting with this approach.
AI tools can be used to support news writing and editing. However, this should always be done under direct editorial supervision. These tools are best used to prepare preliminary drafts. For journalists, AI can help identify trending topics, suggest potential sources, summarize lengthy documents, conduct background checks and even engage audiences more effectively.
While investing in AI infrastructure may place an additional financial burden on media houses in the short term, it could prove vital to their long-term survival. On one hand, AI can significantly enhance the quality and efficiency of news production; on the other, a compact AI-powered newsroom can help reduce human resource costs. Compared to other countries, Nepali media remain behind in adopting technology. But the use of AI in newsrooms is no longer a distant possibility, it is a present-day reality. The question is not whether to use AI, but how to use it effectively to harness its benefits.
