Who decides what becomes news today? Principally, journalists and editors play the central role in deciding which events and issues deserve public attention. But the process is changing now. The news we see, read and engage with is being increasingly shaped by forces beyond the newsroom. This shift raises a question: Who, or what, is now influencing the news production process and overall news agenda?

Algorithms at present influence every stage of news production, distribution and consumption. For example, many newsrooms globally use AI and algorithms to identify potential news stories even before journalists start collecting information. A number of newsrooms use tools like Slack-based newsbots to help journalists identify trending signals and social media mentions at an immense scale. It is interesting to note that with technological advancement, the traditional priority of ‘newsworthiness’ is being ruthlessly outpaced by a data-driven logic of ‘share-worthiness’.

Algorithms do not necessarily prioritise democracy, pluralism or social welfare for news stories; instead, they ask which content will generate the most clicks, shares and watch time. There are also some cases where the news media have started practicing automated journalism. For instance, The Washington Post uses an AI ‘robot reporter’, called Heliograf. Twana Nasih Ahmed and colleagues describe these phenomena as the ‘Click, Code, and Consequences’ framework, where the ‘Click’ economy pressures reporting and the ‘Code’ of AI-driven automation create new ethical dilemmas.

As algorithms gain influence, the role of the editor has also evolved. Their responsibilities extend beyond traditional editorial judgment to include Search Engine Optimisation (SEO), data-driven content decisions and headline optimisation. Hagar and Diakopoulos , in their 2019 article Optimizing Content with A/B Headline Testing: Changing Newsroom Practices, note that many news organisations use A/B headline testing to compare multiple headline versions and identify which performs best with algorithms and audiences.

Likewise, the distribution of news has transitioned from a centralised broadcasting model to a decentralised ‘datacasting’ paradigm. In this environment, the flow of content is determined entirely by algorithmic suggestions and data patterns rather than a broadcaster’s editorial control. This shift has created profound platform dependency. A 2024 Reuters Institute report found that the majority of people across 47 global markets get their news primarily via social media and search engines, with only one-fifth starting their journey on a news website or app.

The algorithmic era has also radically transformed the news consumption pattern. We have moved towards ‘distributed discovery’, where news is not something people intentionally seek out but something they encounter incidentally while browsing their feeds. By repeatedly surfacing content that aligns with a user’s historical data and pre-existing beliefs, these systems can reduce exposure to diverse perspectives and contribute to ideological polarisation. This environment has also contributed to a global rise in ‘news avoidance’, as audiences feel overwhelmed by toxic or emotionally draining content surfaced by engagement-driven metrics.

In this context, the flow of content is organised based on data that the platform collects from users’ interactions. Although it were the editors and the newsroom that decided what and when to show, present and offer, the order of content in the audience’s end, consumed through social media feeds, is now determined by algorithms.

McCombs and Shaw in the Agenda Setting Theory (1972) argue that throughout the history of journalism, it has been up to the editors and newspapers to decide what their audience sees, reads, hears and consequently forms opinions. However, with the advent of digital technology, algorithms influence editorial decisions and control news visibility. Once news gets into social media and other platforms, algorithms dictate what is going to be seen, placed first, recommended, or hidden based on the preferences and interests of each user, not just the news organisations’ criteria.

Algorithms have given rise to what scholars call ‘traffic media’, a business model that aims to maximise and control user attention through algorithmic technologies and corporate policies. At the same time, traditional business models based on advertising and subscriptions have weakened, while technology giants such as Google and Meta have captured a larger share of the digital advertising market. It is important to note that there is growing confrontation over revenue sharing between digital platforms and media organisations. A 2025 study by Peter Chiridza and Admire Mare, titled Digital Platforms and Revenue Generation Strategies Adopted by Zimbabwean Mainstream News Publishers, highlights that most social media platforms do not offer meaningful revenue-sharing partnerships to news publishers, exacerbating these organizations’ economic vulnerabilities.

Early signs of all the shifts described above have gradually emerged in Nepali media ecosystem as well. Some Nepali newsrooms are prioritising news stories based on the degree of audience engagement calculated in terms of clicks, views, shares, likes and other metrics. This has led to an increase in space allotted to lighter genres, such as entertainment, lifestyle, celebrity and highly-shareable content, while neglecting more serious, investigative or public-service journalism. Likewise, journalists in Nepali media are also engaging more directly with social media, including the reporting or re-packaging of issues that have already gone viral. Similarly, at the level of dissemination, newsrooms are increasingly struggling to achieve reach and visibility, as stories that are algorithmically identified, ranked or recommended are more likely to circulate widely.

The way forward is not to resist algorithms but to use them responsibly while preserving journalism’s core values of accuracy, independence, fairness and public interest. In the very spirit, Nepali news organisations should invest in AI and digital literacy so that journalists understand how algorithms shape news visibility and audience engagement. Newsrooms should adopt hybrid models where AI supports tasks such as content discovery, translation, transcription and audience analytics, while journalists retain the editorial decisions.

Nepal can also learn from international practices. The European Union’s Digital Services Act (DSA) has introduced stronger platform accountability by requiring transparency about algorithms and measures to address the spread of misinformation. Australia’s News Media Bargaining Code provides an important example of how governments can reduce the unequal bargaining power between digital platforms and news publishers by requiring technology companies to compensate media organisations for news content. Similarly, Nepali media can adopt transparency initiatives inspired by global efforts such as The Trust Project, NewsGuard and the Journalism Trust Initiative, to learn how to promote greater openness about editorial standards, ownership, funding and sources to strengthen public trust.