For Nepal, the key lesson is that AI governance must evolve alongside technological innovation. Its absence increases the risks of misinformation, deepfakes, algorithmic bias, copyright violations, and declining public trust.
Nepal's media landscape is changing faster than its regulatory framework can keep pace. Nepal has seen a massive surge in digital platforms, with more than 4,600 news sites registered. At the same time, traditional newsrooms are currently in a state of free fall, marked by the shuttering of print editions, stagnant revenue, and a dramatic exodus of professional journalists. Membership in the Federation of Nepali Journalists (FNJ) has also dropped from over 13,000 to around 9,340 following a "purification" campaign, and newspaper sales are falling. Perhaps, the most alarming red flag is the perceived erosion of the primary media watchdog's autonomy.
In this shift, Artificial Intelligence (AI) has quietly entered the newsroom. A significant portion of the workforce already utilises these tools for tasks such as generating headlines and translating stories to summarising interviews and creating images and videos, transcription, and fact-checking. However, lack of formal training and the absence of clear guidelines leave the media industry vulnerable to algorithmic bias and factual inaccuracies. The use of AI has also not been acknowledged and brought under the radar of the Press Council Nepal (PCN), the statutory and autonomous media regulatory body.
Currently, PCN operates under the Press Council Act of 1992, a law written long before the internet transformed how we get news. While the Journalist Code of Conduct (CoC) 2073 (amended 2076), issued by the PCN, stresses factual reporting, it remains almost entirely "silent" on the use and specific ethical risks of AI. Newsrooms are using AI and already experimenting without a safety net.
A research article by Rajiv Timalsina, published in the International Journal of Latest Technology in Engineering, Management & Applied Science, shows that 38% of Nepali journalists use AI for transcription and 22% for fact-checking. Alarmingly, only 12% have received formal training. Without clear guidelines from the PCN, newsrooms are navigating a "regulatory fog" where the line between efficiency and deception is very thin.
Is it really the Council's job to look at AI? The answer is a clear "yes." As the statutory body mandated to uphold professional journalism and public trust, Press Council Nepal should therefore take a proactive role in guiding to develop AI-specific newsroom guidelines and revising the Code of Ethics. Journalism is grounded in the Social Responsibility Theory, which requires the media to serve the public interest through accurate, ethical, and accountable reporting, yet AI's opaque "black-box" decision-making challenges these fundamental principles by reducing transparency and accountability.
While the local regulator in Nepal remains stagnant, the global community is already taking bold steps to manage the AI revolution through new guidelines and declarations. In January 2025, the Press Council of Indonesia issued Regulation No. 1 of 2025, a landmark rule that enforces a "human-machine-human" approach to ensure that a person is always in control of the news from start to finish. Similarly, the Media Council of Kenya developed a comprehensive Media Guide on AI that emphasises human rights frameworks and mandates a "Human-in-the-Loop" principle to protect editorial oversight. In Europe, the Dutch-speaking press council in Belgium was the first to integrate AI guidelines into its code, and the Finnish Council for Mass Media acted as early as 2019 by issuing statements on clearly marking news automation to maintain audience trust. This momentum is growing globally.
These examples demonstrate that if independent media councils do not take the lead in governing this "black box," the freedom of the press may be jeopardised by external or legislative control. Now a curiosity has been generated, what drove them to go for the regulations? The answer is, these councils did not act because AI is perfect, but because it is unpredictable.
A comprehensive examination of 37 AI guidelines across 17 countries by de-Lima-Santos et al. (2024) reveals a global consensus on the "Human-in-the-Loop" (HITL) principle, asserting that AI should augment rather than supplant the unique qualities, creativity, and ethical judgment of human journalists. Similarly, research from the University of Helsinki (Rydenfelt et al., 2026) identifies that journalism's future resilience depends on "human connection," arguing that if newsrooms surrender their core organisational tasks to algorithms, they risk total marginalisation as audiences prioritise trust and authenticity over automated "pink slime" content.
Experiences from countries such as Indonesia, Malaysia, Finland, and members of the European Union show that AI can improve newsroom efficiency, strengthen fact-checking, and expand access to information when supported by clear regulatory frameworks, transparency, and continuous capacity building. Conversely, the absence of robust governance increases the risks of misinformation, deepfakes, algorithmic bias, copyright violations, and declining public trust. For Nepal, the key lesson is that AI governance must evolve alongside technological innovation rather than lag behind it.
Given the rapid emergence of AI in journalism, the PCN should urgently update the Journalistic Code of Conduct to incorporate clear AI ethics standards, require transparent labelling of AI-generated content, institutionalise human editorial oversight, strengthen AI literacy among journalists, and establish ethical guidelines for the procurement and use of AI tools. Equally important, the Council must reinforce its institutional independence to ensure credible, impartial, and effective regulation in the digital age.
Pokharel is a media researcher, poharel.umesh@gmail.com