Abstract
For generations, news media were the chief arbiters of public discourse and the delivery systems of public information to the citizens of democratic societies. Over the past two decades, however, that status quo has drastically changed, as social media platforms became a central conduit for, and functional gatekeeper of, a large portion of news and public knowledge (Bruns 2018; Napoli 2019). Some publishers find themselves in an increasing state of “dependency” with regard to internet, social, and technology platforms—particularly in terms of public attention and web traffic—and have seen their practices and products shaped by asymmetric power relations (Meese and Hurcombe 2021; Nielsen and Ganter 2018). This dependency has damaged journalism in multiple ways, such as the costly and ultimately failed “pivot to video” that many news organizations undertook to chase Facebook’s favor (Madrigal and Meyer 2018; Vaidhyanathan 2018), the collapse of Twitter as a vital community and sourcing tool for journalists (Pate 2023), and pending worries about a decline in web traffic referrals to news sites because of Google’s use of AI Overview summaries (Savage 2025).
The 2024 Digital News Report by the Reuters Institute finds that the majority of people surveyed across 47 global news markets indicated that social media, search, or digital aggregators constituted their main access to news. Only about one-fifth of those surveyed said that news websites or apps are their main way of accessing news. By contrast, YouTube alone is used for news by almost a third of citizens globally (Newman 2024). In the United States, 40 percent of TikTok users, 37 percent of Facebook users, and 30 percent of Instagram users report regularly getting news on those platforms (Shearer et al. 2024). In general, societies are seeing a movement away from
Therefore, within the context of media history, it has fallen to a variety of relatively new commercial information technology firms to establish (or not) novel governance policies with respect to news content, platform users who engage with news, and the professional media outlets that produce news. There has been a great deal of critical scrutiny of the social platform–news publisher relationship, particularly from the perspective of journalism studies (Rashidian et al. 2018). Some journalism advocacy organizations have argued that platforms are openly and systematically hostile to, and destructive of, news media (
To be sure, the erosion of modern journalism represents a complex and dynamic situation that began early in the social media era and cannot be attributed entirely to the rise of competitor platforms. Many of the commercial content-related functions of traditional journalism, such as displaying ads and offering listings for real estate and goods, became unbundled, with large digital platforms replacing news publishers across many potential revenue-generating areas. Journalism institutions failed to innovate at crucial early moments to accommodate new digital and networked realities (Abernathy 2016; Huey et al. 2013; Wihbey 2019). Further, serious internal issues such as long-term declines in news quality, relentless reputational attacks by political elites, and an associated plummeting of public evaluations of news (leading to a growing phenomenon of “news avoidance”) have compounded this situation, making it much more difficult for professional journalism to thrive in the new information environment (Fletcher et al. 2025; Patterson 2013; Villi et al. 2022).
In this article, I argue that as traditional journalism recedes as an anchoring force relevant to mass politics, and as democratic backsliding and threats to democracy increase globally (Haggard and Kaufman 2021; Levitsky and Ziblatt 2018), platforms must rethink the basic structure of their governance models to account for public interest values—namely, the need to promote high-quality news production and protect the creation of democratically important informational goods. The rise of misinformation generated by artificial intelligence (AI) and so-called AI “slop,” or low-quality content that increasingly pervades social media, may threaten the very utility, and experience, of platforms for users (Adami 2024). Indeed, one could easily argue that it is a global imperative that platforms begin to shrink the divide between their modes of operation and those of traditional news publishers, taking measures to help boost and resuscitate professional journalism in new forms.
After reviewing the dynamics of platform governance that affect both individual users engaged with news and news publishers themselves, the latter section of this article advances a new normative model that synthesizes the traditionally conflicting positions of social media and news firms about gatekeeping, curation, and information ethics. The governance model advanced here incorporates a range of ideas emerging from media theory, computational social science, technology policy, and human rights concepts.
The Changing Information Environment
As Nieborg and Poell (2018) have noted, the logic of “platformization” has characterized a new era in the political economy of media and cultural production. Platformization, where content such as news loses its intrinsic value and becomes a “contingent cultural commodity,” has upended the more predictable two-sided market where traditional media served as an intermediary between audiences and advertisers (Nieborg and Poell 2018, 4276). Instead, we now see a complex, volatile, and hierarchical multisided market dominated by a small number of platform corporations.
Although there was an early moment in the platform–publisher relationship when dependency looked more mutually beneficial, recent years have seen nearly all such optimism evaporate. Large platforms like Google and Meta have spent more than a decade now in a highly fraught and antagonistic relationship with news publishers. In recent years, platforms have blocked news content from appearing in some global markets, ended fact-checking arrangements with journalism organizations, shuttered partnerships with news publishers, and generally refused to negotiate over revenue sharing. For example, in 2023, Meta banned news content from appearing in the feeds of Canadian Facebook and Instagram users, in objection to legislation that required Meta to negotiate with news publishers over digital revenue (Parker et al. 2024).
Meanwhile, news outlets have put up more paywalls, making their content functionally inaccessible to social media users who try to access shared links. Publishers sought to strike a better balance between what users want—a clean, information-rich experience—and the imperatives of revenue generation. A demand-side signal for an ad-free experience may have driven the paywall trend in part, but it has resulted in platform users often being unable to access, engage with, and share high-quality content without subscriptions (e.g.,
Even as this has happened, more independent journalists and news-adjacent influencers have created new, grassroots fundraising–focused business models that leverage social media to drive subscriptions and donations (Hurcombe 2024). At the end of the day, platforms have indeed sent publishers enormous amounts of potentially monetizable web traffic, but they have also created a crisis in news production by capturing nearly all advertising revenue. One of the only exceptions to Section 230 of the Communications Decency Act and its legal immunities for online platforms is copyright/intellectual property, and media companies (entertainment and news alike) have often been the most vigorous actors in pursuing takedown requests and enforcement mandates from social media and digital platforms.
Deep and enduring cultural and value conflicts that distinguish publishers and platforms have led to a continual clash over rights, resources, and public attention. Professional news organizations often prioritize public service and holding power to account, while social media platforms focus on serving a global user base with diverse interests; accountability for content moderation is often diffuse among users and platform managers, with platform policy changes and content rules subject to corporate profit interests. These diverging platform–publisher interests drive tensions over how information is created, distributed, and valued in the digital public sphere.
Platforms have effectively won the battle, dwarfing news companies across every conceivable measure of public attention, advertising revenue, and cultural relevance. But a serious challenge looms, relating to basic social order, trust, and stability. From the COVID-19 pandemic beginning in 2020 to national elections around the world in 2024, some major platforms have at times felt compelled to adopt a more journalism-like posture of social responsibility by performing fact-checking and labeling problematic content, downranking misinformation and conspiracy theories, and even changing algorithms to emphasize higher-quality content. In order to operate in many countries, platforms increasingly need to impose further restrictions or modifications: In places where government authorities are likely to impose censorship (e.g., Turkey, Hungary, or India), they may impose restrictions surrounding political topics, and in others (e.g., Brazil and Germany), they may introduce modifications that emphasize prodemocracy information.
Most such editorial decisions by platforms have been temporary, with major platforms having pulled back from more-active interventions in response to political pushback. Many experts maintain that a “crisis of trust and truth” continues unabated (Commission on Information Disorder 2021, 8). As global political trends favor authoritarian populism, civic institutions come under pressure and fail (Zuckerman 2021), and disinformation campaigns damage knowledge institutions (Bennett and Livingston 2020), platforms increasingly reflect the consequences of societal whiplash, pervasive fear, and political instability. Some of these trends are plausibly direct outcomes of platforms’ headlong race to pursue user engagement at any cost (Fisher 2022). Accordingly, platform regulation laws have become increasingly popular, and come into force, across diverse parts of the world—the European Union, the United Kingdom, Brazil, India, Australia, and Thailand, among others. Around the world, leaders increasingly blame social media for societal ills and regularly threaten companies with legal action and even the arrest of employees.
There has long been evidence that news media consumption may be correlated with political trust, although the relationship may vary over time (Strömbäck et al. 2016). Today, trust in news may be receding across many global societies as information environments become less structured by traditional news access, especially television news (Fletcher et al. 2025). Meanwhile, news access on social media has been linked to lowered trust in news (S. Park and J. Lee 2023).
This historic movement away from direct news access to distributed discovery on digital platforms is producing what Nielsen and Fletcher (2020, 139) have called “democratic creative destruction,” with uncertain consequences for open societies and democratic processes. As professional news media and journalism recede in power and lose their role as anchors of the information environment that can generate trust and a sense of shared reality, the pressure on social platforms to fulfill that role is likely to mount. While the future in this regard is unclear, now is perhaps the proper moment to reflect on how platforms might respond affirmatively to address the deep structural problems emerging in the information environment.
Such reflection will be especially important with the emergence of generative AI platforms. With AI playing a new kind of intermediary role between internet users and information (i.e., summarizing news articles and related sources of information and serving them to users directly), we are likely to see an even more contentious platform–publisher relationship. This dynamic is already raising concerns about the fate of high-quality journalism within an AI-dominated context and stoking fears about news institutions’ further loss of governance capacity, control over their outputs, and ability to remain financially viable. Many companies with social platforms are also developing AI models (for example, Meta and Llama, Google/YouTube and Gemini), and social media and generative AI are likely to become intertwined in myriad ways relating to content creation, moderation, and governance.
Governance and News Users
Platform logics can exert a subtle influence on the global infrastructure of public expression, creating incentives and disincentives as to what types of digital content gain validation. News media are not governed necessarily by platforms but rather
Broadly speaking, platform regulatory models can be conceived within a tripartite framework: 1) self-governance, where platform firms write their own rules; 2) external governance, where governments establish oversight; or 3) cogovernance, which involves civil society organizations helping to formulate and encourage platform standards and often to exert pressure toward adoption of these standards (Gorwa 2019). Platforms typically have community standards or terms of service that limit the types of content that can be shared or produced on them. Typically, these rules may restrict hate speech, nudity, and other types of content that other users may find offensive or threatening. At times, these rules can affect what kinds of news-related activity users may engage in (particularly with regard to more extreme types of news). But on the whole, few restrictions curb, ex ante, discussion of mainstream, professional news stories, news reporting activity, or user sharing or commentary.
Yet there are nuances within platform standards, and the stringency of rules may vary substantially—ranging from the wide-open norms found on Telegram and X to the (relatively) more rules-bound, circumscribed environments of YouTube and Instagram. In terms of platform content moderation, more decentralized platforms such as Reddit may offload substantial governance power to community moderators, and others (often smaller platforms) may have more “artisanal” or case-by-case content governance rules and structures (Caplan 2018). Newer innovations and interventions in the domain of information communications technologies, such as the rise of federated platforms and ideas for user-centric middleware (software allowing more individual control over data and information environments), may expand traditional governance frameworks to empower users and user communities further (Keller 2021; Kleppmann et al. 2024).
Napoli (2015) has noted that the traditional public interest concepts that structured and governed the regulation of broadcast news media in the twentieth century have morphed in the social media era in two distinct ways. First, updated notions of public interest focus primarily on restrictive frameworks (i.e., governing what should not be posted, shared, or surfaced) and neglect affirmative visions for news and information. These platform frameworks often build on the “deterrence model common in legal settings” (Venkatesh et al. 2021, 3). Second, this era has also seen the advance of increasingly individualist frameworks for media governance: “Many of the responsibilities associated with the production and dissemination of the news and information essential to a well-functioning democracy fall, within the context of social media platforms, to individual media users, who are presumably provided with platforms that enable them to fully exercise their abilities to access and disseminate news and information” (Napoli 2015, 757).
Algorithmic governance
Beyond governance of individual users and their news-related behaviors, wider dimensions relate to the governance of information dissemination and exposure as well as algorithm-based amplification levers. In the parlance of the social media world, we might think of this as the governance of virality and “reach”—the computer code that guides which posts are shown to many persons. Algorithmic dynamics in this space can be organized into several categories of power: connective or network standards; information quality; collaborative filtering; and user dissemination mechanisms.
Digital and social platforms have a shaping function that not only mediates what news audiences see but also creates distinct incentive structures and constraints on news outlets and journalists themselves. In terms of connective or network standard-setting, Nielsen and Ganter (2022) note that platforms exert myriad forms of power that affect news publishers and the members of their organizations. These powers include platforms’ ability to set standards for users and communities; the capacity to “make and break connections” within networks by changing technical protocols; and the use of information asymmetry (i.e., privileged data collection and processing) to structure online environments, target content, and perform automated transactions at scale (Nielsen and Ganter 2022, 21).
With regard to the issue of algorithms and information quality, the scholarly discourse on governance has increasingly focused on those algorithms that curate content for users in their feeds. Of central interest to social scientists has been the degree to which quality information versus low-grade or even harmful misinformation might be algorithmically surfaced and promoted/amplified. Much of the debate over algorithm governance and public interest considerations has centered on false, biased, or fake “news” of various kinds (Grinberg et al. 2019) and on the general kinds of news that users typically share about issues such as elections, public health, and politics. The relative degree of exposure to ideologically biased news has also been a subject of intensive study (Budak et al. 2024; González-Bailón et al. 2023). Scholars have found that platform algorithms frequently homogenize the news content most often shown to users, thus limiting the diversity of news displayed and sometimes overrepresenting U.S. outlets such as Fox News (Nechushtai et al. 2024).
Further, there is the question of how collective signals by users shape the behavior of algorithms, creating a cyclical and at times self-reinforcing pattern. Traditional media theory has focused on the concept of gatekeeping, which relates to the ways that media institutions act to control, select, and organize public news and information. Contemporary scholars have updated the traditional model of control to account for the ways that the aggregation of actions (e.g., sharing, liking, reposting) on decentralized platforms produce gatekeeping power (Wallace 2018). Computer scientists have described these platform dynamics as “collaborative filtering,” whereby the digital signals produced by users interact with recommendation algorithms to produce an information and news environment that distills and ranks items from the vast sea of potentially visible content. This highly technical domain of collaborative filtering, which seeks to structure and make meaning of innumerable digital signals, continues to be an area of focus among machine-learning and AI researchers (Kleinberg et al. 2024; Wu et al. 2022).
Users’ abilities to engage with and disseminate news is not in fact uniform. In terms of restrictions, governance of individual users on platforms also involves algorithmic downranking, which is the deprecation or suppression of user-generated content based on certain criteria that are not typically available to users. Platforms may decide to make news content (and other types of content) largely invisible when posted by users. Debates over such “shadow-banning” techniques have led to wide-ranging political controversies and spurred regulatory moves across the world (Leerssen 2023; Langvardt and Rozenshtein 2025).
How algorithms understand users and their desires through the collection of data—how they may deliver the automated, sometimes uncanny, fulfillment of user desires and gratifications, or what Lazer et al. (2024) call “agency hacking”—is now a central question to be evaluated for potential downstream effects on society and citizenries. Whether it is the treatment of a user’s posts or the amplification or downranking treatment of shared news, platforms are governing the news-related user experience across many dimensions and at nearly every possible level.
Platform Affordances and News Effects
Although there are many types of media producers now operating on social platforms, the governance structures that affect professional news outlets and their working journalists remain of central importance. How do platforms govern news outlets and journalists on their platform? The answer may vary by platform, but generally speaking, platform managers make choices with significant governance-related consequences.
The governance of news producers might be described and evaluated at five levels: technical, cultural-environmental, normative, monetization, and commercial. These overlapping areas of influence affect both how people consume news and how news outlets and journalists produce it:
There are also deeper exploratory questions about the cultural effects of social media as conditioning factors in how people come to expect information to be delivered. These questions go beyond the scope of what we might think of as traditional platform governance. Yet, as platforms themselves often become the public sphere, particularly in countries with less developed media ecosystems, they take on new powers. By being nearly coextensive with the entire media ecosystem, they effectively establish the norms and default expectations for media content. They are, in effect, rewriting news standards. Such dynamics hearken back to the classic insights of contemporary media theory, stemming from the likes of Marshall McLuhan, Lewis Mumford, and others who theorized how medium and message—technology and media content—are radically intertwined in modern societies (Carey 1981).
For example, the meteoric rise of short-form video (e.g., YouTube Shorts, Instagram Reels, and of course, TikTok) has created new expectations among audiences for how information should be conveyed in a compelling visual fashion and how claims about public issues and the very concept of facticity are typically conveyed (i.e., through assertion, not attribution and sourcing, and through highly compressed visual sequences that leave out context). Not only are social media companies serving as de facto governors over news stories in the public sphere and over their potential public audiences, but the platforms are restructuring global citizenries—psychologically and politically—in their fundamental concept of news. They condition the public, recalibrating definitions and rewiring preferences and cultural expectations (Waisbord 2018; Wihbey 2019).
News depreciation
If news is effectively a depreciating asset in democratic life, to what degree have social media firms and their governance policies been, on balance, responsible—both inadvertently through their meteoric success and intentionally through their competitive/aggressive behavior? How much are platforms destructive of and deleterious to the very concept of high-quality news?
It is well known that there has been a collapse of journalism’s business model as tech companies have drawn away advertising dollars. Professional news media have contributed to negative judgments and loss of trust through an emphasis on conflict, negative news, and sensationalism (Patterson 2013). Further, ideological attacks have accelerated the loss of public trust in the profession (McChesney 2004; Smith 2010).
Yet other factors have played their part in the depreciation of news. Findings from the social sciences help to contextualize some of these contributing factors, including the nature and incentives around human content-sharing behavior and the extent of algorithmic bias. In particular, four areas have substantially changed how the public sees and interacts with news: toxicity, negativity, and news avoidance; institutional reputational damage and attacks on journalists; information quality confusion; and substitution effects. Setting aside questions about actual malicious intent by platforms, or more causal questions about the degree of direct responsibility, it is worth scanning the broad field of evidence about the net effects of platform operations on news media.
First, how have social media platforms contributed to public perceptions of toxicity with regard to news? Such sentiments appear to be driven in part by the environment of social platforms (Mathews et al. 2024; Watson et al. 2024). Negativity in general has a greater association with news engagement in online environments (Robertson et al. 2023). Related to these problems is the global phenomenon of “news avoidance,” a trend that has been on the rise across global societies and one likely driven in part by social media algorithms (Villi et al. 2022).
How have social media platforms contributed to sentiments about journalists and journalism organizations, particularly as they frequently provide relatively ungoverned avenues for motivated persons and groups to attack organizations and individuals (Tandoc et al. 2023)? Journalists are frequently targets of propaganda and disinformation campaigns in online spaces; computational strategies by malicious actors, sometimes using bots and automated systems, seek to manipulate journalists, create false impressions about the veracity of information they provide, or intimidate or threaten them (Woolley 2023). It remains an open question how much platforms see (or should see) journalists as special actors who might deserve enhanced protections, or whether they should normatively be treated just as any other user.
Further, how have content moderation regimes contributed to general confusion about news quality and undermined the authority of professionally produced news? All too often allowing misinformation to run rampant, they create an atmosphere of epistemic chaos and doubt among users (Udry and Barber 2024). The resulting atmosphere of skepticism and confusion often sees professional, high-quality news placed into a kind of value-neutral content blender, where audiences lack quality signals on the platform interface.
Relatedly, how are we to think about what economists might call “substitution effects”—that is, the replacement of higher-priced goods (in this case, higher-quality news) by lower-priced ones (news with lower civic relevance or quality [Jang and M. Park 2016; Toff and Mathews 2024])? While it may be unfair to fault a platform for serving up what the masses of users may want, the design of social platforms and the stickiness, even addictive qualities, of their personalized algorithms, may produce a central externality relating to high-quality news about public affairs and events—namely the impression that it is more boring, less desirable, and less personally compelling or relevant than the endless stream of more entertainment-driven, personalized content.
Different factors drive substitution effects. Although content moderation contributes, the biggest factor is likely to be algorithms that prioritize engagement metrics (e.g., high numbers of likes, shares, comments). Such algorithms select for surprising or highly emotional content, which is more likely in low-quality sources because the high-quality ones focus on facts (which change slowly, for the most part). Other factors driving substitution effects include the low cost of publicly circulated content (relative to content purchased through direct subscriptions) and the speed of (superficial) access to news through quick headlines and overview visuals on the platforms.
Taken together, the changes wrought by the social media companies on the health and production of democratically valuable news content are substantial, and they have wide implications for the future of democratic societies. To formulate better governance around news content will require that the product teams at social media companies more carefully consider the downstream effects of platform design decisions as they relate to news and information integrity and, ultimately audience trust.
Networks and News
The degree of contraction of professional journalism and the sidelining of news media within the public sphere varies from country to country. However, the trendlines collectively point to a period when the institutions of journalism will eventually be scarce in many countries, with little power left to inform and engage broad swaths of the public. Globally, there are numerous efforts to try to stop this destruction and revitalize journalism, from civil society actors and philanthropies to calls for public media and nonprofit news creation.
It is not clear that the platforms will seek to support journalism in any particular way, but I suggest that platforms will eventually need publishers of professional, factual news, and that they will need journalists. The reason for this is that platforms are businesses rooted in political cultures, and as such, they will become increasingly vulnerable to hostile governments and citizenries as democracies and open societies come under more strain and face greater challenges. I do not argue that platforms will
Through the twentieth century, pre-internet, many societies had just a few television news stations and newspapers in a given media market, and the choices for citizens were severely constrained. Much of the critical media studies literature focused on news monopolies and corporate journalistic power—until social media arrived (Bagdikian 2004; McChesney 2004).
The research literature about what are termed “high-choice” media environments—where there are myriad options for accessing media—is reasonably clear about the consequences for citizens in democracies. Prior (2007) analyzed early “post-broadcast democracy,” wherein high-choice environments result in significant knowledge “inequalities.” Only those people who are highly motivated to pursue political information (often partisans) actively locate and digest it. Meanwhile, most citizens gravitate toward various forms of entertainment—now abundantly available via online platforms—intensifying asymmetries in political awareness and engagement and making elections more polarized.
Algorithms, which have the capacity to exacerbate extant social and political differences, make the situation all the more complex and potentially problematic. Greater partisan polarization in societies such as the United States has become entangled with algorithmic personalization and tailoring of content on platforms, amplifying “selective exposure” effects (Stroud 2011) and creating circular patterns wherein: 1) partisans are attracted to more attitude-conforming content; 2) the algorithms adjust to these preferences; and 3) platform users ultimately see more extreme, partisan, and opinion-based media content. Social algorithms (Lazer 2015) then, in effect, may work against the very logic of neutral, professional reportorial journalism, which seeks to inform wide audiences based on shared realities and mainstream cultures of interpretation.
While the diversity and abundance of news and information may be seen as a theoretical virtue—breaking former monopolistic media strangleholds, providing more avenues for marginalized voices—the consequent gaps in citizen knowledge amid high-choice media environments present new problems. The situation in which many societies find themselves has prompted media observers to contemplate the proper contours of a more balanced and stable system: What does an effectively governed “news ecosystem” look like? How would we know this exists or has come into being?
In the context of U.S. media and politics, Ladd (2012, 7) articulated a view that there could be some more ideal point between a hierarchical and highly flat, decentralized ecosystem, what he called: A balance between a highly trusted, homogeneous media establishment with little viable competition and an extremely fragmented media environment without any widely trusted information sources. In this middle path, individuals would continue to have a wide range of choices, including partisan, sensational, or conventionally objective news, as well as the option of avoiding news altogether. Yet the remaining institutional journalists and news outlets would continue to transmit important political information, with a significant portion of the public retaining enough confidence in the institutional press to use this information to hold government accountable.
Ladd’s idea of an ideal point between the top-down and bottom-up paradigms comes from a reflection on a highly developed, media-rich country. Yet as the United States and other economically advanced and politically mature countries in the West see increasing problems with fundamental democratic processes, it is worth examining the literature on what human rights scholars refer to as “institutionalized media freedom theory,” which specifies the crucial elements of effective systems that guard fundamental freedoms and value individuals’ rights against the state and other powerful societal actors.
Comparative study of media environments indicates that mere increases in freedoms of the press, speech, and expression do not ensure social order or good governance. Rather, as Snyder (2022, 149) has noted, these freedoms need to be embedded in a “reasonably well-institutionalized setting.” Such a setting is characterized by journalists exercising independent judgment and working within strong, independent media organizations and also by political institutions allowing people to act on information and ideas circulating in the public sphere. Snyder also notes, crucially, such settings generally require the condition that “the products of quality journalism are more readily accessible than degraded political speech” (Snyder 2022, 149).
The idea that quality journalism is essential to democracy does not necessitate the maintenance of some kind of ivory-tower or corporate news media. The institutionalized media freedom theory idea should be qualified and elaborated upon, especially given how valuable social media and networked communications have been to journalists and news organizations in reporting on social problems (van der Haak et al. 2012; Wihbey 2019). Quality journalism in the digitally networked era is often produced and, in effect, cocreated by a combination of professional and amateur actors, with experts, social critics, and interested citizens often helping to shape stories and media narratives.
Outside of the news and platform ecosystem, there are constitutive actors, such as citizen activist organizations and government agencies, news influencers and content creators, whose voices and perspectives can enhance, contribute to, and frame quality journalism. In the digitally connected, social media–driven environment of today, journalists can seek to produce “networks of recognition,” with news serving an aggregation and catalytic function within the public sphere so that diverse actors can see how they are mutually connected, reflect on those connections, and build shared public knowledge (Wihbey 2019). Such a pattern often involves journalists connecting disparate persons by building stories and sharing them through networks, facilitating citizen dialogue and political action, and bridging “structural holes” (Burt 2004; Wihbey 2019, 42).
Platforms and publishers may seem, ostensibly, to represent two cultures—even two competing (largely tacit) theories of democracy. Traditional journalism wants an informed public, facilitated by news gatekeepers (facticity, curation). By contrast, social media want an open public sphere, where individuals seek what interests them and are driven by individual motivations (voice, choice) and are free to connect. Both centrally seek public attention. But what the idea of networks of recognition does is acknowledge the need to blend the hierarchical and networked visions. Indeed, informed publics must be networked publics for collective action to be possible and for democracy to function; networked publics that are not also informed publics are unlikely to contribute to truly democratic outcomes (and are susceptible to demagoguery and authoritarian populism).
Such a networked journalism theory potentially helps alleviate some of the negative consequences/externalities of high-choice media environments (e.g., knowledge asymmetries), while also concretizing Ladd’s “middle path” vision of a news ecosystem that embodies an ideal blend of both institutionalized newsmakers and networked information offerings. Further, such networked but professionalized approaches mitigate the potential charge of elitism that comes with reifying institutional news media, even as institutions—not just platforms—retain an essential place in the public sphere.
To return to the question of how platforms might come to see publishers, their values, and their work as integral and essential, a theory of networked journalism might be substantially aided by a new approach to algorithms and information design that focuses on optimizing for reporting on social problems and civic issues. Helping citizens, civic actors, and news media to connect, find one another, and collaborate in building knowledge could be a more intentional goal of social platforms. How to build richer communities in online environments remains a key area of focus by researchers (Forestal 2022).
Protections and supports: A global imperative?
More research has begun to focus on both the rationale and the mechanics for producing generally higher-quality social media feeds curated by recommendation algorithms that take wider account of user signals and optimize for more than just engagement or time on site—the short-term engagement metrics for which platforms have typically optimized. One proposed metric is known as “long-term user value,” defined as “outcomes that align with individual users’ deliberative, forward-looking preferences or aspirations. Long-term user value prioritizes long-run user preferences and users’ ability to achieve their aspirations over short-run, impulsive preferences” (Moehring et al. 2025, 4). Such new approaches might involve algorithms taking greater account of “nonengagement” signals produced by platform users and scanning for more subtle signals that help retain users over longer time periods—not just optimizing for engagement metrics such as comments, likes, and shares, for example, on highly emotional, extreme, or frivolous viral content (Cunningham et al. 2024). Social media companies’ corresponding research focusing on product design, particularly on algorithms, means more critical thinking about upstream choices that structure algorithmic and machine-learning systems by specifying better objectives within computer code—and less of a focus on content moderation policy and enforcement (Iyer 2025).
How exactly changing algorithms and product design to maximize longer-term user value would affect news content is not immediately clear, but surely high-quality, rigorously fact-checked content that has a public interest dimension could be given greater consideration in ranking and recommendation algorithms than such content currently is given. There is evidence that Meta has invested in and used a “news ecosystem quality” measure to boost higher-quality content at crucial times, such as elections (Jackson 2024).
Beyond the need for algorithmic restructuring, platforms need to focus on several other governance-related areas in order to support news capacity and protect the services that journalists provide—both for society and, indeed, the social media companies themselves. These include: 1) support for the consistent exercise of freedom of the press rights, which in many ways platforms themselves benefit from (and will face increasing damage without such rights); 2) protecting journalists from manipulation, attacks, and threats, so they can produce high-value, factual content; and 3) establishing and renewing broad, albeit at arm’s length, financial supports for journalism, potentially allowing journalists and news media to participate more deeply in the creator economy of platforms.
Let us first consider how important press freedoms are for platforms, which typically benefit from publisher-like rights in many global jurisdictions. It is often underappreciated how much social media companies’ ability to operate relatively wide-open free-speech spaces (i.e., an open platform for user-generated content) essentially depends on, perhaps exploits, the historical, hard-won freedoms earned by the professional press over generations. Consider the landmark
In a general sense, social media platforms have enjoyed
The issue of journalist safety on social media is also a pressing issue, and one for which platforms have many capabilities and tools they could deploy to help. Maria Ressa, the Nobel laureate, has called attention to the harassment that many journalists face—particularly female journalists—and the lack of response when platforms hear such complaints. Ressa’s case has highlighted the need for a concerted effort by digital platforms, civil society, and government to ensure that journalists can meaningfully exercise their rights without fear (Posetti 2017; Posetti et al. 2021). In global surveys, women journalists regularly report higher rates of threats and abuse online, which can force them to pull back from their work and even consider quitting the profession (Oukhiar 2021; Stahel and Schoen 2020). Woolley (2023) has noted that bot attacks (use of computationally driven networks of accounts) often target journalists in an effort to push them off platforms or limit their engagement. To help combat such attacks globally, platforms would need to devote more resources (including recruiting more subject matter experts and native speakers) to monitor and address such patterns of intimidation and threats.
Finally, the collapse of the business model of journalism has particularly decimated accountability, investigative reporting, and local news functions (Rolnik et al. 2019). Public service-oriented journalists now struggle to find jobs or make a living. Meanwhile, with the monetization of content and labor on digital platforms, disparities grow as “news influencers” and content creators find ways to make money on platforms; but professional, institutional journalists are not afforded such opportunities. Content creators are sometimes given revenue shares on platforms such as YouTube (if their audience is sizable enough); on some platforms, they are given special tools to help them build stories and audiences. These evolving relationships are complex and not without problems (Hödl and Myrach 2023).
While paying such independent entrepreneurs for their time seems inherently fair and positive—it has created a whole new economy of sorts—it does highlight the degree to which institutional, public service journalism has, for the most part, been left out. Journalists spend enormous amounts of time on social media platforms, facilitating dialogue, engaging audiences, informing persons, and generally improving the empirical environment. Almost none of this individual professional work is compensated. Of course, large platforms point out that they drive platform users to news websites and help journalists and news outlets build valuable brands. Yet as problems deepen in journalism—and, as this article has suggested, it behooves platforms to rethink their strategy with regard to news—it may be that some kinds of monetary supports are justified, and the precedent is now well established with influencers and creators.
Conclusion
The divide between traditional news media and social platforms is a consequential structural transformation of the global information environment. As this article has documented, platforms now serve as primary conduits for news consumption across many global markets, yet their governance models remain largely misaligned with the public interest values that have traditionally anchored professional journalism. The governance dynamics affecting both individual users and news publishers on these platforms—spanning technical, normative, cultural-environmental, and commercial dimensions—have contributed to the depreciation of high-quality news through mechanisms that include increased toxicity, institutional reputational damage, information quality confusion, and substitution effects. These changes have coincided with, and perhaps partly driven, democratic backsliding worldwide, raising urgent questions about whether platforms must adopt more journalism-like postures of social responsibility.
Further, with the rise of generative AI, we are seeing the emergence of a sector that is selling “answer machines” that require high-quality inputs—which not infrequently means news content. In some ways, the demand for high-quality news is as high or higher than ever, yet the distribution and economic models for news provision are breaking or already broken.
The path forward lies both in greater and more targeted research and serious policy and strategy shifts by platforms. There is a pressing need for a research agenda that more systematically draws connections between the platform governance of news and the downstream democratic consequences for the health of the public sphere (Lazer et al. 2024). Further, as generative AI technologies begin to add new complexities and layers to this dynamic news–social space, scholars must articulate more clearly the consequences of governance decisions and point to alternative pathways that may lead to more prosocial outcomes and democratic health. Models for monetization of content and data stand to change in the coming AI era, with the fate of high-quality news media—and the prospects for informed democratic citizenries globally—hanging in the balance.
The future relationship between platforms and publishers may ultimately be determined less by current antagonisms than by emerging global imperatives. As professional journalism recedes in power and capacity to influence the public sphere, platforms may well face increasing pressure from hostile governments and disillusioned citizenries. This scenario suggests that platforms might eventually recognize that supporting professional, factual news serves their long-term strategic interests. A model that synthesizes the hierarchical vision of traditional journalism with the networked paradigm of social media could help mitigate the negative externalities of high-choice media environments while preserving the essential functions of both institutions. Such an approach would require algorithmic restructuring to optimize for long-term user value rather than mere engagement, enhanced protections for press freedom and journalist safety, and renewed financial support for news production. Of course, there is no guarantee that social media platforms, which are often backed by large global corporations, will necessarily recognize the growing tensions and opportunities this article identifies.
The emerging governance model proposed in this article incorporates elements from media theory, computational social science, technology policy, and human rights concepts to articulate a path forward. Rather than viewing platforms and publishers as representing fundamentally opposed cultures or embodying totally diverging theories of democracy, this framework recognizes their potential complementarity: Informed publics must be networked publics for democracy to function effectively, while networked publics must be informed to resist demagoguery and authoritarianism. There are broad benefits for platform companies if they begin thinking in these longer-term, more strategic ways. As generative AI introduces new complexities to this dynamic information space, the imperative for platforms to help bridge the news and social media divide grows more urgent. The health of democratic societies worldwide may depend on whether social platforms can evolve their governance models to help resuscitate professional journalism in new forms that serve the genuine public interest within our increasingly networked public sphere.