Social media platforms’ responses to COVID-19-related mis- and disinformation: the insufficiency of self-governance

Warnke, Lina; Maier, Anna-Lena; Gilbert, Dirk Ulrich Article — Published Version Social media platforms’ responses to COVID-19-related misand disinformation: the insufficiency of selfgovernance Journal of Management and Governance Provided in Cooperation with: Springer Nature Suggested Citation: Warnke, Lina; Maier, Anna-Lena; Gilbert, Dirk Ulrich (2024) : Social media platforms’ responses to COVID-19-related misand disinformation: the insufficiency of selfgovernance, Journal of Management and Governance, ISSN 1572-963X, Springer US, New York, NY, Vol. 28, Iss. 4, pp. 1079-1115, https://doi.org/10.1007/s10997-023-09694-5 This Version is available at: https://hdl.handle.net/10419/315326 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) Journal of Management and Governance (2024) 28:1079–1115 https://doi.org/10.1007/s10997-023-09694-5 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑ anddisinformation: theinsufficiency ofself‑governance LinaWarnke1 · Anna‑LenaMaier2 · DirkUlrichGilbert3 Accepted: 11 December 2023 / Published online: 1 February 2024 © The Author(s) 2024 Abstract The spread of misand disinformation on social media platforms is a significant societal threat. During the COVID-19 pandemic, misand disinformation played an important role in counteracting public health efforts. In this article, we explore how the three most relevant social media platforms, Facebook, YouTube, and Twitter, design their (IT) self-governance as a response to COVID-19-related misand disinformation, and provide explanations for the limited scope of their responses. Exploring the under-researched connection between the operating principles of social media platforms and their limited measures against misand disinformation, we address a relevant research gap in the extant literature on digital platforms and self-governance, particularly the role of IT governance (ITG), providing the ground for our argument against an overreliance on self-governance. In our qualitative study that draws on publicly available documents, we find that the shortcomings of current responses to misand disinformation are partly due to the complex nature of misand disinformation, as well as the wider political and societal implications of determining online content’s factuality. The core problem, however, is grounded in the current overreliance on self-governance. We argue for an enhanced dialogue and collaboration between social media platforms and their relevant stakeholders, especially governments. We contribute to the growing ITG literature and debate about platforms’ roles and responsibilities, supporting the intensifying calls for governmental regulation. Keywords COVID-19· Disinformation· IT governance· Misinformation· Selfgovernance· Social media platforms 1 Introduction Digital platforms assume increasingly powerful roles in society (Lindman et al., 2023). The rise of social media platforms as a subtype of digital platforms has been accompanied by increasingly critical accounts of their destructive potential Extended author information available on the last page of the article 1080 L.Warnke et al. 1 3 (Cusumano etal., 2022), as the heated debates in connection with Twitter’s acquisition by Tech billionaire Elon Musk illustrate (BBC, 2022). The platform has since been renamed “X”, however in this paper we will refer to it as Twitter.The reliance on social media as younger generations’ primary information source and the uptake of misand disinformation on these platforms increase this threat (Marin, 2021, p. 2; Reuters Institute for the Study of Journalism, 2021a). This was particularly observable during the Covid-19 pandemic, which has also been referred to as an infodemic, prompting scholarly calls for measures against both misand disinformation and their underlying causes (Marin, 2020). The term infodemic means “a flood of information on the Covid-19 pandemic”, which has been fueled by misinformation and disinformation spreading on social media (World Health Organization [WHO], 2021b). How social media platforms manage and govern misand disinformation can be understood as a matter of both ethics and governance. Recent research has advocated for exploring how Information Technology Governance (ITG) may be used to proactively address ethical issues related to different kinds of information technology (IT) (Wilkin & Chenhall, 2020). With this article, we contribute to the growing ITG literature by exploring how the three most relevant social media platforms, i.e., Facebook, YouTube, and Twitter, design their self-governance measures to respond to COVID-19-related misand disinformation and provide explanations for the limited scope of these responses. Self-governance in this context is defined as regulations and guidelines that are issued by either a single company or a group of companies (industry self-governance) and applied to themselves to manage and control their businesses (Cusumano etal., 2021). The case of COVID-19-related misand disinformation on social media platforms is especially useful for this study as it has gained immense attention from various stakeholders on a global scale. Further, it demonstrated the destructive potential of misand disinformation spreading on social media. Roozenbeek etal., (2020, p. 12) find that a higher susceptibility to misinformation is directly linked to people’s behaviors during the pandemic, resulting in vaccine hesitancy and less adherence to public health measures. Therefore, the COVID-19-pandemic presents a valuable empirical context for understanding social media platforms’ self-governance mechanisms. Corporate governance as the study of how an organization is governed and how decisions are made is a critical element in analyzing social media platforms’ responses to misand disinformation. ITG is described as “an integral part of corporate governance and addresses the definition and implementation” of three key aspects: governance structures, processes, and relational mechanisms, which enable “both business and IT people to execute their responsibilities in support of business/ IT alignment and the creation of business value from IT-enabled business investments” (Van Grembergen & De Haes, 2009, p. 3). We therefore understand ITG as a form of self-governance. IT governance structures refer to “organizational units and roles responsible for making IT decisions and for enabling contacts between business and IT management (decision-making) functions” (Van Grembergen & De Haes, 2009, p. 21). ITG processes are designed to ensure the alignment of daily business routines to corporate policies and provide feedback through “the formalization and institutionalization of strategic IT decision-making or IT monitoring 1081 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… procedures” (Van Grembergen & De Haes, 2009, p. 22). Lastly, relational mechanisms are announcements, channels, and educational efforts that are designed through participation and collaboration between the different corporate levels consisting of executives as well as business and IT managers. The latter is especially important for business and IT alignment (Van Grembergen & De Haes, 2009, p. 22). We contribute to extant ITG research by exploring governance challenges linked to social media platforms as particularly relevant and powerful actors in the context of COVID-19-related misand disinformation. We specifically focus on social media platforms which “enable people […] to make connections by sharing expressive and communicative content, building professional careers, and enjoying online social lives” (van Dijck, 2013, p. 4). Globally, there are 4.8 billion active social media users (DataReportal etal., 2023) with an average daily use of 2 h and 22 min (Buchholz, 2022). Therefore, social media platforms are a large part of most people’s lives and, hence, may significantly impact them and society more generally. A rather new and unintentional function of social media platforms is gathering information and news. A study by the Reuters Institute for the Study of Journalism (2021a) showed that 67% of respondents under the age of 25 use social media as a source of information. At the same time, relying on social media as the main information source results in more susceptibility to misinformation (Allcott & Gentzkow, 2017, p.17). Against this background, misinformation and disinformation have become particularly salient and consequential. Vosoughi etal., (2018, p.1146) define misinformation as “information that is inaccurate or misleading”. The degree of intention differentiates misinformation from disinformation. While misinformation is unintentional, disinformation spreads false stories deliberately (Geeng etal., 2020, p.1). The COVID-19 pandemic has not been the first event that led to an increasing spread of misinformation and disinformation on social media platforms. The 2016 U.S. elections are a rather prominent example underlining the societal and political relevance of misand disinformation on social media platforms (Allcott etal., 2019). The extent to which false information is spread deliberately is hard to assess. The three social media platforms we study correspondingly seem to prefer to speak of misinformation when describing their measures. However, most phenomena they refer to in this context are, in fact, better described as disinformation, for example, anti-vaccination conspiracy theories. Effectively countering disinformation on social media platforms has become particularly relevant in the ongoing COVID-19 pandemic. Disinformation that spreads quickly and widely undermines confidence in public health measurees and thus negatively impacts crisis management (Algan etal., 2022). Lack of trust in governments and limited scientific knowledge contribute to the consumption and spread of misinformation and disinformation on social media (Chowdhury et al., 2021). This is particularly problematic as anti-vaccine content and corresponding disinformation efforts have been found to directly contribute to vaccination refusal (Muric etal., 2021), thus counteracting public health efforts. Our main argument, and at the same time our central contribution, is that the spread of such harmful information is not accidental, but rather part of the fundamental design of social media platforms. It can thus be argued to be grounded 1082 L.Warnke et al. 1 3 in their basic operating principles and practices of value creation. To understand how a platform’s value is created, it is essential to understand the basic structure of the underlying business model. Van Dijck etal. (2018, p.10) point out that digital platforms monetize attention, data, and user valuation, which arguably affects their processes, governance structures and relational mechanisms, as well as measures to moderate and manage content. Zuboff (2015, 2019, 2022) argues that this is problematic because users of social media platforms are hardly aware of their role, thus creating a “behavioral surplus” that, in turn, creates revenue for the social media platforms, incentivizing the latter to manipulate users’ online behavior so as to create further revenue. The kind of information a user sees while using social media is highly determined by the platform’s algorithms. Users can become trapped in filter bubbles as search results are personalized and preselected based on personal characteristics such as location or previous searches (Kompetenzzentrum Öffentliche IT, 2016, p.99). Because algorithms are trained to promote provoking, sensational content that users engage with, and misinformation fits many of these criteria, they are often spread through the platforms’ algorithms (Avaaz, 2020; Culliford, 2020; Eisenstat, 2021; Roberts, 2020). Consequently, this sparks a debate concerning the societal and political roles and responsibilities of social media platforms. Recent attempts to regulate social media platforms more strictly such as the Digital Services Act have been unsuccessful (Turillazzi etal., 2023), and, thus, in the absence of sufficient government regulation, an (over)reliance on self-governance by social media platforms themselves is observed. Cusumano etal., (2021, p. 1273f.) argue that self-regulatory responses of many platforms were adopted too late and remain insufficient to address current challenges. Empirically exploring the under-researched connection between governance principles and measures against misand disinformation, we thus ask: How do social media platforms design their (IT) self-governance as a response to COVID-19-related misinformation and disinformation? To answer this question, this article proceeds as follows. We discuss some central operating principles of social media platforms, ITG mechanisms, their interaction and lastly address the current overreliance on self-governance in contrast to government regulation. In the subsequent methods section, we justify our case selection and provide more granular details on the empirical context of our study. We then present our core findings regarding the nature and scope of strategic responses of three social media platforms of interest. We then discuss our findings, highlighting potential pathways towards rebalancing voluntary action in terms of IT self-governance and government regulation. Our conclusion contains the most relevant contributions and addresses some of the limitations of this study. 2 Interdisciplinary literature review In the following, we describe the most relevant features of social media platforms that lead to governance problems in general and misor disinformation in particular, both of which present important and critical issues to be addressed through effective self-governance in general and ITG in particular. It is thus important to first 1083 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… understand the basic operating principles of social media platforms and how they generate governance problems before clarifying the relationship between self-governance and government regulation. 2.1 Operating principles ofsocial media platforms asafacilitator ofmis‑ anddisinformation Digital platforms create value through network effects, which according to Parker etal., (2016, p. 17) “refers to the impact that the number of users of a platform has on the value created for each user”. According to Srnicek (2017, p. 256), network effects are possibly the most value-defining feature of a platform that draw in more and more users and eventually lead to a monopoly. Algorithms, algorithmic decision-making, and the corresponding network effects are central to the functioning of digital platforms in general and social media platforms in particular (Zarsky, 2016). In the context of social media platforms, they (co-)determine what content users see and engage with. Srnicek (2017, p.254) argues that data are central to social media platforms and the main source of their economic and political power. Through users’ information, photos, and activities, social media platforms are provided with more and more data, which teaches their self-learning software to better understand and predict users’ actions (Royakkers etal., 2018, p.139). Therefore, the cost of their power is the privacy of users. However, in contrast to other challenging areas of digitalization, data protection and privacy are receiving the most legal attention and supervision, for example, provided by the European Data Protection Regulation. Still, critics doubt the effectiveness and suitability of these regulations. Although there are some laws in place already, they do not necessarily consider the intertwined relationships between data, users, algorithms, and other factors that contribute to the functioning of platforms (European Commission, 2022a). Personalization, as one of the main features determining why users are attracted and loyal to a particular platform (van Dijck etal., 2018, p.42), creates a lock-in effect. The algorithms governing this process are kept secret as accurate predictions generate a competitive advantage. The algorithms on which personalization is based result in search engines being biased. Pariser (2011, p.9) refers to this as filter bubbles, “unique universes of information” created by engineers that affect how ideas and information are perceived. Through the creation of individual online universes, the personalization movement is also threatening democracy (Zuboff, 2022). Democracy relies on taking on different viewpoints and shared facts, which are increasingly undermined by individuals’ personalized environments online. Further, it limits one’s autonomy and severely impacts meaningful decisions as the possible options presented were preselected by algorithms (Pariser, 2011, p.16). Similarly, echo chambers are an increasingly central issue in the spread of misinformation and disinformation on the Internet. Echo chambers refer to communities, especially on social media platforms, that share the same worldview (Colleoni etal., 2014, p.319). Established views are reinforced within these echo chambers, and members are rarely exposed to alternative views and opinions (Lütjen, 2016, p.17). 1084 L.Warnke et al. 1 3 Disand misinformation in line with a certain echo chamber’s view and ideology are diffused more quickly through the echo chamber (Törnberg, 2018). In contrast to filter bubbles controlled by algorithms, echo chambers develop through personal action. Moreover, social media platforms consciously and by design encourage predominantly negative emotional content, e.g., expressions of fear and anger (Steinert, 2020). In sum, filter bubbles and echo chambers are best understood as a direct result of a social media platform’s basic operating principles and form a central part of their value creation. At the same time, governance is costly, and implementing governance measures often creates tensions between economic value creation and governance costs (Huber etal., 2017). Therefore, implementing and potentially extending governance that would, for example, target filter bubbles and echo chambers on social media platforms, might be seen as a direct threat to the business models of large social media platforms relying on network effects. 2.2 IT governance Misand disinformation, we argue, constitute a societal threat that is facilitated through social media platforms and reinforced through their basic operating, business model-informing principles. This is enabled through several circumstances. According to Marin (2021), sharing is a split-second decision regardless of its truth content. Further, Geeng etal. (2020) find that most users do not investigate the content they are sharing. This contributes to false information spreading faster and farther on social media than the truth (Vosoughi etal., 2018, p.1149) and shapes users’ opinions. However, the spread of misinformation is not just promoted by users but also by bots and algorithms. This is where ITG, as discussed earlier, comes into play due to the large role IT plays in governing and shaping social media platforms. As briefly explained in the introduction, ITG broadly refers to governance structures, processes, and relational mechanisms within the respective organization (Van Grembergen & De Haes, 2009). The main task of ITG is to effectively and efficiently enable the organization to create business value, to mitigate risks associated with IT, and to facilitate the alignment between corporate vision, management practices, and the IT infrastructure (Bowen etal., 2007). Most ITG research focuses on the role of ITG for business and IT alignment and the resulting performance effects (e.g., De Haes & Van Grembergen, 2017). How IT Governance mechanisms are used in an organization, among other things, depends on their inherent dependence on IT. Based on Nolan and McFarlan (2005), Héroux and Fortin (2014) categorize an organization’s dependence on IT by means of four IT modes. The IT modes range from highly defensive, somewhat offensive, moderately offensive to highly offensive. Attributes determining the IT mode are IT intensity as well as size and decentralization of the IT function. We argue that due to their business model and as born digitals (Monaghan etal., 2020), social media platforms can be categorized as highly offensive with high IT intensity, a large IT function, and a low-moderate decentralization of their IT function. In highly offensive IT modes, ITG processes and relational capabilities are used to a moderate to high 1085 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… degree, whereas ITG structures are only used to a low to moderate degree (Héroux & Fortin, 2014, p. 161). To gain a better understanding of ITG mechanisms, Héroux and Fortin (2014) use several items to measure each construct in their survey. ITG structures are, for example, specific IT committees for security, projects, or architecture, with the board of directors having both expertise of IT risks and management functions responsible for IT security, risks, and compliance. This understanding is in line with Altemimi and Zakaria’s (2015) identified drivers of ITG structures, which are authority and membership as well as coordination mechanisms, demonstrating that ITG structures determine responsibility and decision-making-functions. Héroux and Fortin (2014) further analyze ITG processes, expressed through formal processes regarding IT strategy, and work with external agencies to conduct IT security audits. Drivers of ITG processes are performance monitoring and the alignment of IT decisions with key business, thus addressing corporate strategy (Altemimi & Zakaria, 2015). Lastly, ITG relational capabilities entail senior executives being involved in shaping a vision and IT’s role in the organization as well as the implementation of the vision throughout the organization (Héroux & Fortin, 2014). This is related to a number of drivers such as leadership, skills, collaborative relationships, as well as role, responsibility, and commitment and thus leads to establishing commitment and support both at the top management levels and throughout the whole organization (Altemimi & Zakaria, 2015). 2.3 ITG andsocial media platforms’ responses tomis‑ anddisinformation There are different approaches to principles guiding self-governance efforts of social media platforms. Marin (2021), for example, established a hierarchy of norms of relevance to social media platforms that also affect the spread of (mis-)information. The first layer consists of legal norms. These are kept to a minimum by the platforms and communicated through terms and conditions, often concerning illegal activities, including hate speech and personal attacks. The second layer concerns meta-norms of sociality that aim to promote the further growth of the network within the law. Lastly, the third layer is the core layer concerning local and unpredictable norms. These depend on the group level and can apply to a community within a particular social media platform and, therefore, differ across different communities using the same platform. Social media platforms usually do not intervene or establish any restrictions to this level as long as these community norms follow the law and contribute to expanding the network (Marin, 2021). This hierarchy of norms addresses the requirement of IT decisions being aligned with the key business objectives and principles constituting ITG processes (Altemimi & Zakaria, 2015). According to Marin (2021), measures to fight the spread of misor disinformation entail both elements of human supervision and algorithms. There are numerous approaches to limiting misor disinformation on social media platforms. Lazer etal., (2018, p.1095) distinguish between two types of interventions. First, approaches that empower users to evaluate the encountered information and make informed decisions about their truth content and whether or not to share it. Second, intervention approaches that involve structural changes to prevent users from 1086 L.Warnke et al. 1 3 coming into contact with disinformation in the first place. Intervention approaches are mostly implemented voluntarily by social media platforms as a key self-governing mechanism in their fight against misinformation and disinformation, or driven by government regulation. Since disinformation can also be considered a central part of a social media platform’s business model, Trittin-Ulbrich etal., (2021, p.15) point out that platforms’ priority would be to avoid or circumvent governmental regulations. For example, collaboration with fact-checkers is promoted. Lazer etal., (2018, p.1096) suggest altering the algorithms to emphasize highquality information and provide information regarding the source’s quality. Furthermore, the personalization of political information should be reduced. These selfgovernance mechanisms could thus potentially majorly interfere with the platforms’ operating principles. Geeng etal., (2020, p.2) analyze Facebook and Twitter and report that both platforms remove inauthentic and manipulative accounts. Moreover, users can manually flag posts, or these may be automatically detected and then demoted. Lastly, Facebook provides more information to users about an article’s source. Before sharing information that is known to be false, the platform warns the user and offers related fact-checked articles (Geeng etal., 2020). Pennycook etal. (2020b) found that nudges, such as accuracy reminders, promote more thoughtful sharing behavior, which would benefit the fight against misinformation that is often shared unintentionally. Especially when it comes to changing platform algorithms to adequately address misinformation and disinformation, ITG processes are needed that are performance monitoring oriented, take into account current developments, and can be adapted to strategic changes at short notice. (Altemimi & Zakaria, 2015). A mechanism available on almost every social media platform is flagging, sometimes referred to as reporting. This practice is defined by Crawford and Gillespie (2016, p.411) as “reporting offensive content to a social media platform”. In this sense, the users are engaged in shaping the platform’s content and, in a way, community values. As flagging is often just the first step in potentially removing content from the platform, this practice adds legitimization to the platform’s final decision (Crawford &Gillespie, 2016, p.412). However, there are also downsides to the practice of flagging. One of the greatest challenges of this mechanism is that users may abuse it. For example, users may flag content as a joke because of an existing feud or competition with other creators, or may even bully and harass creators. Due to the limited communication and interaction, the difference is almost impossible to detect for the platforms, which may leave the mechanism invaluable (Crawford &Gillespie, 2016, p.420). Moreover, Pennycook etal. (2020a) highlight that the approach may lead to an Implied Truth Effect. This states that content that is not labeled is granted higher credibility and is assumed to be accurate, thus creating a false sense of security. Another option is the voluntary collaboration between governments and social media platforms. For example, in the UK, the national health service (NHS) worked together with social media platforms to promote accurate COVID-19-related information and, at the same time, fight the spread of misinformation. Measures included, among others, the verification of governmental accounts to establish trusted sources, direct links, or easy access to accurate information provided by the NHS for COVID19-related searches and the exclusion or removal of identified false information and 1093 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… Table 3 (continued) First order codes Second order themes Aggregate dimensions Cross-sector collaboration Partnerships (Health) Experts Media industry Industry peers Global and national health and governmental organizations Fact-checkers Organizations National Adaptation Community guidelines Community Guidelines Approval in groups by admins Consequences for violating users Lock / remove account Strike system 1094 L.Warnke et al. 1 3 Table 4 Illustrative examples of the data analysis Source Quote First order code Second order theme Rosen (2020) [Facebook] “During the month of April, we put warning labels on about 50 million pieces of content related to COVID-19 on Facebook, based on around 7500 articles by our independent fact-checking partners.” Labeling Empowering user intervention Fact-checkers Partnerships Twitter (2021a) “Our systems learn from past decisions by our review teams, so over time, the technology is able to help us rank content or challenge accounts automatically. For content that requires additional context, such as misleading information around COVID19, our teams will continue to review those reports manually.” Automated flagging Algorithmic screening Manual review Manual screening Twitter Safety (2020) “Starting in early 2021, we may label or place a warning on Tweets that advance unsubstantiated rumors, disputed claims, as well as incomplete or out-of-context information about vaccines. Tweets that are labeled under this expanded guidance may link to authoritative public health information or the Twitter Rules to provide people with additional context and authoritative information about COVID-19.” Labeling Empowering user intervention Jin (2020) [Facebook] “To further limit the spread of misinformation, this week we are launching a dedicated section of the COVID-19 Information Center called Facts about COVID-19. It will de-bunk common myths that have been identified by the World Health Organization such as drinking bleach will prevent the coronavirus or that taking hydroxychloroquine can prevent COVID-19.” Central information centers Empowering user intervention YouTube (2021d) “For content where accuracy and authoritativeness are key, including news, politics, medical, and scientific information, we use machine learning systems that prioritize information from authoritative sources in search results and recommendations.” Machine learning Algorithmic Screening Promoting trusted sources and information Empowering user intervention Improving search results Higher exposure to trusted content / improve recommendation and ranking Structural intervention 1095 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… Table 4 (continued) Source Quote First order code Second order theme Twitter Safety (2020) “Starting next week, we will prioritize the removal of the most harmful misleading information, and during the coming weeks, begin to label Tweets that contain potentially misleading information about the vaccines.” Label Empowering user intervention Remove content Structural intervention Twitter Safety (2021) “The Q&A featured Dr. Anthony Fauci, US President Biden’s chief medical advisor, and other members of the White House COVID-19 response team. In India, we worked with the Ministry of Health to organize Vaccine Vartha, a weekly expert talk hosted on Twitter that enables vaccine experts to answer citizen questions.” Promoting trusted sources and information Empowering user intervention (Health) experts Partnerships National adaptation YouTube (2021a) “Note: YouTube’s policies on COVID-19 are subject to change in response to changes to global or local health authorities’ guidance on the virus.” Community guidelines Rosen (2020) [Facebook] “We are also requiring some admins for groups with admins or members who have violated our COVID-19 policies to temporarily approve all posts within their group.” Approval in groups by admins Consequences for users 1096 L.Warnke et al. 1 3 Approaches and initiatives were mixed, and the data analysis identified common themes across all three platforms. Most analyzed documents specifically address COVID-19 responses. In the following findings section, the identified mechanisms addressing misand disinformation on social media platforms are presented, categorized according to the identified data structure. Understanding and categorizing these responses, although triggered by the specific event of the COVID-19 pandemic, promises to provide the empirical basis for being able to assess future ITG responses to similar large-scale events with considerable societal impact. 4.1 Screening methods Before content can be removed or labeled, it needs to be detected and reviewed. For that purpose, social media platforms apply several different screening methods to identify and evaluate content. Three screening methods can be distinguished, i.e., algorithmic screening, manual screening, and a mixed screening approach. Algorithmic screening includes the use of algorithms to detect false or misleading content. All three platforms apply this method. The platforms claim a high success rate for algorithms detecting content automatically rather than it being reported by users or through manual screening (Facebook, 2021g). The platforms also highlight the advantages of algorithmic screening and the areas in which it is particularly useful. In a report, YouTube (2021g) points out that automatic detection allows for faster and more precise action when enforcing its policies. Further, their machine learning tools are improving in different languages, and fact-checking agencies work in more than 60 languages (Twitter Safety, 2021). The screening process can be divided into two steps. Firstly, content that may be violating the platform’s policies needs to be detected. Secondly, it is reviewed and evaluated before deciding whether it violates policies and which intervention approach should be applied. These two steps may be carried out by either algorithmic or manual screening or a combination thereof. For example, Facebook applies artificial intelligence (AI) to remove COVID-19-related misinformation after the questionable content has been flagged through manual screening (Rosen, 2021). A key characteristic of machine learning is that it needs to be trained by manual inputs. Manual screening includes the screening by platform employees as well as user reporting. This mechanism is still heavily applied by all platforms. Besides relying on their own personnel that manually screens content, during the COVID-19 pandemic, the social media platforms extensively collaborated with partners such as health experts and (governmental) health organizations. The mechanism is not just used in the context of misinformation, but in all violations of the platforms’ community guidelines and policies (YouTube, 2019). Interestingly, Twitter is the only platform that does not communicate about user reporting, but rather communicates that it only relies on screening through trusted partners like public health authorities, NGOs, or governments (Twitter, 2021b). Section4.3 contains more findings regarding partnerships that were implemented as a response to COVID-19-related misand disinformation. 1097 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… According to YouTube, algorithmic screening is the most important and successful method. In second and third place are user reporting and detection through the Trusted Flagger Program. Only a small fraction of the removed videos is detected by NGOs and government agencies (YouTube, 2021e). After questionable content is flagged, it may be reviewed through manual screening as well (YouTube, 2021f). Facebook also cooperates with fact-checkers to make qualified decisions on the accurateness of COVID-19-related content (Rosen, 2021). Although machine learning processes are applied widely across the platform, YouTube (2019) acknowledges that “human expertise is still a critical component of [their] enforcement efforts”. Lastly, a mixed approach combining both algorithmic and manual screening can be found. This may refer to circumstances where algorithmic screening is not yet advanced enough so that, in consequence, manual involvement is still required. Further, a combined approach may be applied after content is flagged automatically. When a decision on further actions cannot be reached by the algorithm, manual screening is required. At Facebook, a mixed approach allows machine learning tools to evolve and be trained to be more effective and efficient (Rosen, 2021). Twitter equally relies on a mixed methods approach and holds that accounts will not permanently be suspended solely based on automated enforcement systems, but only after human review (Twitter, 2021a). After screening content and potentially identifying misleading information, intervention approaches are applied, which will be addressed in the following section. 4.2 Intervention approaches Regarding the question how misand disinformation is dealt with, two types of intervention approaches of self-governance can be distinguished, which have also been introduced by Lazer etal., (2018, p.1095). Platforms choose either an empowering user intervention approach or a structural intervention approach. Empowering users refers to providing tools to support users in making informed decisions. For example, a tool often used by social media platforms is labeling. Thus, content that is known to include false information is not removed, instead a disclaimer is added. Labels are applied either as an explicit warning or by providing links to additional information to offer context to the questionable content. By providing reliable sources, the users are supported in informing themselves about COVID-19 (Twitter Safety, 2020). This approach is applied by all three platforms and often includes links to authoritative sources and third-party sites since a central part of the empowering user intervention is to promote trusted sources and information. In this regard, platforms are also working on improving search results for users who use social media platforms to find credible information about COVID-19 (Twitter, 2021a). In this context, Facebook even created a specialized information center concerning COVID-19, which features real-time updates from organizations such as the WHO (Clegg, 2020). The information center thus addresses various aspects around COVID-19 and collects all relevant information in one place for Facebook’s users. It also aims to educate users of the globally practiced physical and 1098 L.Warnke et al. 1 3 social distancing approaches, guide people with a potential infection, and, lastly, provide links to relevant health authorities and organizations (Facebook, 2021e). The information center is continuously updated and extended, now also including a section addressing misinformation. Twitter created a comparable central place to collect information regarding COVID-19. On the platform, it is referred to as the COVID-19 Events page and “is available at the top of the Home timeline for everyone in 30 + countries” (Twitter, 2021a). Overall, all governance mechanisms pursue the aim to empower users to make their own decisions based on reliable information. A common concept for all three platforms was that empowering user intervention approaches aim to promote informed decision-making (Facebook, 2021g). Facebook is also promoting the development and improvement of users’ news literacy, which is an important skill in the fight against misand disinformation in the long term. Lastly, the platforms may send out messages and alerts to users. In particular, Facebook sends messages to users who have interacted with content in the past, which since has been declared to include misand disinformation (Rosen, 2020). Regarding structural intervention, two approaches are mostly applied, namely removing content or reducing the visibility of content. Facebook clearly states what the conditions are for either approach to be applied. Facebook removes misinformation that is a potential threat to physical integrity. With this step, the platform relies on external health experts such as the WHO. Only content which promotes debunked information is removed (Facebook, 2021b). On YouTube, content that is not an imminent threat is not removed, but rather the visibility is decreased to limit its spread (YouTube, 2021b). This approach is often combined with labeling, an empowering user intervention tool. Content that does not clearly contradict the platform’s guidelines but may be misleading or false is labeled accordingly, or additional contextual information is provided (Twitter, 2021a). In combination with the provision of context and authoritative sources on questionable content, the platforms also pursue an approach where trusted content is given higher exposure and is promoted more through the platform’s recommendation or ranking systems (YouTube, 2021d). By downranking misinformation and highlighting trusted content, the platforms try to foster an environment where users encounter less misand disinformation and are presented with reliable information regarding COVID-19. In addition, fake accounts that only pursue spreading disinformation are being targeted and removed (Mosseri, 2017). This is also particularly pursued by restricting users’ engagement options with content that is misleading but not removable. For example, Twitter (2021b) disabled engagement functions while content is being reviewed. The platforms actively pursue a combination of these two intervention approaches and their different tools. This shows that, on one hand, machine learning mechanisms can be applied to stop misinformation from spreading and to reduce the exposure of users to misinformation. Nevertheless, on the other hand, the social media platforms acknowledge that it is also important to educate users and improve their news literacy skills in order for them to make informed decisions. In the long run, this might also improve the platform’s problem of being polluted by mis-and disinformation. Although both approaches are actively pursued and combined by the 1099 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… platforms, our analysis showed that structural intervention is pursued much less frequently than empowering user intervention. 4.3 Partnerships In many of their intervention approaches and several screening methods, the platforms cooperate with external partners. The most mentioned form of partnering is with health organizations or health experts. This often occurs in combination with “promoting trusted sources and information” and “labeling”, which are both empowering user interventions. All platforms pursue the aim to provide users with authoritative information. For that purpose, the platforms rely on public health experts, public health organizations, as well as on governments. While the most mentioned global health organization is the WHO, a focus is set on national adaptation as well. The platforms also provide links to national health organizations, often alongside global information from the WHO, and adapt their mechanisms to local circumstances (Twitter, 2021a). The contribution of the WHO includes the provision of links to their website to provide a trusted and reliable source for users where they could find further information about the virus without the threat of encountering misor disinformation. Further, the WHO also publishes common COVID-19-related myths and “debunked” these on their website, which provides the social media platforms with a baseline and reference in regard to which content is inaccurate and serves as a guide for decision-making (Rosen, 2020). In addition, Facebook tasks fact-checkers with reviewing content. Over time, this practice has expanded so that the platform now works with over 80 independent fact-checkers, allowing content to be reviewed in over 60 languages (Rosen, 2021). To ensure the independence and quality of the fact-checking organizations, they “are certified through the non-partisan International Fact-Checking Network, which is a subsidiary of the journalism research organization “The Poynter Institute” (Facebook, 2021f). Another large group of partners includes experts. For example, the platforms show that they frequently consult global health experts when developing new strategies and policies. Twitter even helps experts to be heard and found by verifying their accounts, which might increase their reach and credibility (Twitter Safety, 2020). In this area, Twitter also organizes events where health experts can interact with users and answer questions concerning the virus (Twitter Safety, 2021). Less common partnerships include the cooperation with industry peers, organizations, and the media industry. The cooperation with organizations mostly aims to support users’ news literacy (Facebook, 2021f). The cooperation with the media industry aimed to support and protect journalists to ensure the availability of qualitative and reliable information and was mostly accomplished through donations (Facebook, 2021h; Twitter, 2021a). The platforms mention “[w]orking together with industry peers to keep people safe” (Twitter, 2021a), but concrete actions are not communicated. 1100 L.Warnke et al. 1 3 Lastly, the platforms also promote cross-sector collaboration working together with the aforementioned groups as well as with users, governments, and NGOs. As an example, we can cite YouTube’s Trusted Flagger Program that we described earlier or a similar approach adopted by Twitter (2021c). 4.4 Community guidelines andconsequences forviolations Community guidelines build the basis of the platform governance and are established based on core values such as freedom of expression (Facebook, 2021c). They are also a key element in the platform’s fight against misand disinformation. The guidelines contain policies and rules as to what is allowed and what is prohibited from being posted on social media platforms. Although the exact content of the community guidelines and the type of content that is prohibited go beyond the scope of this article, it is interesting to note that the guidelines specifically identify COVID-19-related topics that are subject to consequences. Thus, rather than formulating vague guidelines which may provide more leeway for both users and platforms, the platforms have decided to implement very specific policies. As new conspiracy theories or myths regarding COVID-19 develop and are debunked by official sources, these need to be added to the guidelines so that the platforms are able to limit the spread of such content (YouTube, 2021a). YouTube (2019) also provides more insights into the development process of community guidelines and their adaptations, showing that the platforms are eager to involve various stakeholders to improve their service. Further, they have noted that their COVID-19 policies “are subject to change in response to changes to global or local health authorities’ guidance on the virus” (YouTube, 2021a). Of equal concern are the consequences for users or content that violates the platform’s community guidelines. This is in part a complement to the intervention approaches already mentioned. There are different consequences or corrective measures for content and users who violate the platforms’ rules. For example, Twitter (2021b) established a strike system where different types of violations lead to the user accumulating strike points. The more strikes a user accumulates, the more severe the consequences are. After a 12-h account suspension, a 7-day suspension is imposed. If a user accumulates more than five strikes, the account is locked permanently. By this measure, the platform intends to foster a learning effect by increasing users’ awareness of policies (Twitter Safety, 2021). YouTube and Facebook use similar strike systems. Another mechanism introduced by Facebook requires group admins to temporarily review and approve group content for groups whose members have previously violated COVID-19 policies (Rosen, 2020). Group admins are also responsible when the content they approved contains a violation, and may receive a strike (Facebook, 2021b). 5 Discussion Social media platforms pursue various approaches of self-governance to counter misand disinformation related to COVID-19. In the following, we discuss our findings and use the mechanisms described in the ITG literature to answer our 1101 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… research question of how social media platforms design their IT self-governance as a response to Covid-19-related misand disinformation. The platforms’ responses identified in our analysis can be related to the current ITG framework consisting of structures, processes, and relational mechanisms as summarized in Table5. In this sense, screening methods and community guidelines can be understood in terms of ITG processes. Whether content violates the platform’s rules is decided through formal procedures for IT-related decision-making that are executed through the screening methods and community guidelines. Further, they facilitate and enable the interaction between management and business operations. Structural intervention can also be understood in terms of ITG processes. The decision on limiting content’s exposure through technical interference constitutes both the decision-making aspect and responsibility of ITG structures as well as the establishment of an IT strategy and policy as ITG processes entail. Empowering user intervention can be interpreted in light of ITG processes as an implementation of the platforms’ IT strategies and policies to educate users. Further, the identified consequences for users who violate the platforms’ are part of the formalized ITG processes, which are based on community guidelines and thus align business routines to corporate policies. Lastly, the identified partnerships can be interpreted in the light of ITG relational mechanisms focusing on the partnerships’ collaborative characteristic. At the same time, external partnerships present an addition to the current ITG framework as it is not solely limited to collaboration within the organization but also includes external partners. We propose that in other ITG-related situations, involving external partners in the ITG mechanisms would also benefit the organization. As indicated earlier, ITG structures do not emerge from our extensive body of data. This absence already suggests that creating new ITG structures by, for example, setting up IT committees with specific expertise, would contribute to the effectiveness of social media platforms’ responses to misand disinformation. The data show that, although both empowering user intervention and structural intervention approaches are widely applied across all three platforms, structural Table 5 Social media platforms’ responses to misand disinformation related to ITG mechanisms ITG mechanisms Social media platforms’ responses to misand disinformation Structures No findings in our data Processes Screening methods Algorithmic screening Manual screening Intervention approaches Structural intervention Empowering user intervention Consequences for violations Community guidelines Relational mechanisms Partnerships 1102 L.Warnke et al. 1 3 intervention approaches were less present in the data. Since structural intervention, e.g. altering algorithms to prioritize verified content when adapting rating and recommendation systems, has a greater impact on the platforms’ business model, this in turn potentially increases the costs of governance activities. It appears that platforms are more reluctant to apply and develop those intervention approaches. Our study thus empirically contributes to the growing literature on the interlinkages of platforms’ operating principles and their ITG efforts. The broader ITG literature supports this interpretation, as it shows that governance activities are generally associated with considerable costs, which, in turn, impacts the design and extent of such activities (Huber etal., 2017). Going beyond the general role of (self-)governance costs, other studies have shown that content containing misor disinformation is attracting more attention to social media platforms than other content (Vosoughi etal., 2018, p. 1149) and is thus ranked higher by the algorithms (Avaaz, 2020; Culliford, 2020; Eisenstat, 2021; Roberts, 2020). This is supported by our observation of empowering user intervention approaches being pursued over structural intervention approaches which would actively interfere with the platform’s algorithms and might alienate some users. These are further indicators that the platforms’ economic interests in growing the platform might dominate and affect the scope of their interventions, which would require active alterations to their systems such as those presented in our findings. Going back to Marin’s (2021) conceptualization of a hierarchy of norms determining the spread of unintentional and deliberate false information, we see that through the core layer, platforms enable echo chamber forming and problems such as misinformation spreading. By keeping legal norms in the first layer to a minimum, there is greater room for platforms to evade their responsibility without having a compliance issue. Thus, we argue that the norms should be revised and that platforms should place more emphasis on the first layer. By increasing the legal norms, a safe environment can be established and dominating problems such as misand disinformation can be reduced. Focusing their responses on empowering user intervention approaches allows platforms to defer their responsibility to users without actively decreasing the amount of misand disinformation on their platform. This is further supported by van Dijck etal., (2018, p.147), who state that platforms and their stakeholders “need to put long-term public value creation over short-term economic gain”. Furthermore, such measures would indicate a shift from voluntary self-governance to compliance with legal norms, i.e., government regulation, thus potentially reducing complexity. When governance challenges related to managing content on social media platforms become a compliance issue, we argue, this may result in more responsible processes (e.g., daily business routines), governance structures (e.g., roles and responsibilities), and relational mechanisms (e.g., collaboration between different corporate levels) given a reduced level of ambiguity. At the same time, initial investments into corresponding governance measures might eventually lead to a reduction of governance costs (Huber etal., 2017). ITG research should establish whether and to what extent measures on the spectrum of government regulation (compliance) and self-governance (voluntary action) result in deor increased complexity or costs. 1109 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… relationship between business model and governance measures beyond conceptual theorization. Funding Open Access funding enabled and organized by Projekt DEAL. No funding was received for conducting this study. Data availability The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors have no relevant financial or non-financial interests to disclose. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/. References Algan, Y., Cohen, D., & Péron, M. (2022). Why is trust key to managing crises? World Economic Forum. Retrieved March 12, 2022, from https:// www. wefor um. org/ agenda/ 2022/ 02/ trustfacto rscovid 19crisis/ Allcott, H., Gentzkow, M., & Yu, C. (2019). Trends in the diffusion of misinformation on social media. Research & Politics, 6(2), 1–8. Allcott, H., & Gentzkow, M. (2017). Social media and fake news in the 2016 election (NBER Working Paper Series). Cambridge, Massachusetts. Alphabet. (2022). Form 10-K: annual report pursuant to section13 or 15(d) of the securities exchange act of 1934 for the fiscal year ended December 31, 2021. Retrieved September 15, 2022, from https:// abc. xyz/ inves tor/ static/ pdf/ 20220 202_ alpha bet_ 10K. pdf? cache= fc816 90 Altemimi, M., & Zakaria, M. (2015). Developing factors for effective IT governance mechanism. In 2015 9th Malaysian Software Engineering Conference (MySEC), (pp. 245–251). Avaaz (2020). Facebook’s algorithm: A major threat to public health. Avaaz. Retrieved September 10, 2022, from https:// secure. avaaz. org/ campa ign/ en/ faceb ook_ threat_ health/ Bartley, T. (2007). Institutional emergence in an era of globalization: The rise of transnational private regulation of labor and environmental conditions. American Journal of Sociology, 113(2), 297– 351. https:// doi. org/ 10. 1086/ 518871 BBC. (2021). Social media: should people be allowed to be anonymous online? Retrieved September 10, 2022, from https:// www. bbc. co. uk/ newsr ound/ 56114 122 BBC. (2022). Elon Musk warned he must protect Twitter users. Retrieved September 10, 2022, from https:// www. bbc. com/ news/ busin ess61225 355 Bowen, P. L., Cheung, M. Y. D., & Rohde, F. H. (2007). Enhancing IT governance practices: A model and case study of an organization’s efforts. International Journal of Accounting Information Systems, 8, 191–221. https:// doi. org/ 10. 1016/j. accinf. 2007. 07. 002 Briggs, M. (2020). Assessment of the code of practice on disinformation. MediaWrites. Retrieved May 22, 2022, from https:// media writes. law/ asses smentofthecodeofpract iceondisin forma tion/ 1110 L.Warnke et al. 1 3 Buchholz, K. (2022). Where people spend the most & least time on social media. In Statista. Retrieved May 22, 2022, from https:// www. stati sta. com/ chart/ 18983/ timespentonsocialmedia/ Chase, P. H. (2019). The EU code of practice on disinformation: the difficulty of regulating a nebulous problem. Transatlantic Working Group Working Paper. Chowdhury, N., Khalid, A., & Turin, T. C. (2021). Understanding misinformation infodemic during public health emergencies due to large-scale disease outbreaks: A rapid review. Journal of Public Health. https:// doi. org/ 10. 1007/ s1038902101565-3 Clegg, N. (2020). Combating COVID-19 misinformation across our apps. Meta. Retrieved May 22, 2022, from https:// about. fb. com/ news/ 2020/ 03/ comba tingCOVID19misin forma tion/ Cohen, M., & Sundararajan, A. (2015). Self-regulation and innovation in the peer-to-peer sharing economy. University of Chicago Law Review Online, 82(1), 116–133. Colleoni, E., Rozza, A., & Arvidsson, A. (2014). Echo chamber or public sphere? Predicting political orientation and measuring political homophily in twitter using big data. Journal of Communication, 64(2), 317–332. https:// doi. org/ 10. 1111/ jcom. 12084 Crawford, K., & Gillespie, T. (2016). What is a flag for? Social media reporting tools and the vocabulary of complaint. New Media & Society, 18(3), 410–428. https:// doi. org/ 10. 1177/ 14614 44814 543163 Culliford, E. (2020). On Facebook, health-misinformation ‘superspreaders’ rack up billions of views: report. Reuters. Retrieved May 11, 2022, from https:// www. reute rs. com/ artic le/ ushealthcoron avirusfaceb ookidUSK CN25F 1M4 Cusumano, M. A., Gawer, A., & Yoffie, D. B. (2021). Can self-regulation save digital platforms? Industrial and Corporate Change, 30(5), 1259–1285. https:// doi. org/ 10. 1093/ icc/ dtab0 52 Cusumano, M. A., Yoffie, D. B. & Gawer, A. (2022). Pushing social media platforms to self-regulate. The Regulatory Review. University of Pennsylvania Law School. Retrieved February 14, 2023, from https:// www. t here grevi ew. org/ 2022/ 01/ 03/ cusum anoyoffiegawerpushi ngsocialmediaselfregul ate/ Daniel, E. (2021). Twitter introduces “strike system” for vaccine misinformation. Verdict. Retrieved April 11, 2022, from https:// www. verdi ct. co. uk/ twitt erstrikesystem/ DataReportal, Meltwater & We Are Social. (2023). Number of internet and social media users worldwide as of April 2023 (in billions). InStatista. Retrieved June 20, 2023, from https:// www. stati sta. com/ stati stics/ 617136/ digit alpopul ationworld wide/ De Haes, S., & Van Grembergen, W. (2017). An exploratory study into IT governance implementations and its impact on business/IT alignment. Information Systems Management, 26(2), 123–137. https:// doi. org/ 10. 1080/ 10580 53090 27947 86 EFRAG. (2022). European sustainability reporting standard SEC1 sector classification standard: Working paper. Retrieved May 11, 2022, from https:// www. efrag. org/ News/ Proje ct572/ EFRAGpubli shestodaythenextsetofPTFESRSClust erWorki ngPapers Eisenstat, Y. (2021). How to hold social media accountable for undermining democracy. Harvard Business Review. Retrieved June 11, 2022, from https:// hbr. org/ 2021/ 01/ howtoholdsocialmediaaccou ntableforunder miningdemoc racy European Commission. (2022a). Questions and answers Digital Markets Act. Retrieved October 9, 2023, from https:// ec. europa. eu/ commi ssion/ press corner/ api/ files/ docum ent/ print/ en/ qanda_ 20_ 2349/ QANDA_ 20_ 2349_ EN. pdf European Commission. (2022b). Tackling online disinformation. Retrieved January 18, 2024, from https:// digit alstrat egy. ec. europa. eu/ en/ polic ies/ onlinedisin forma tion European Parliament. (2021). Social media and democracy: We need laws, not platform guidelines. Retrieved May 10, 2022, from https:// www. europ arl. europa. eu/ news/ en/ headl ines/ socie ty/ 20210 204ST O97129/ socialmediaanddemoc racyweneedlawsnotplatf ormguide lines Facebook. (2021a). Company info. Retrieved June 20, 2022, from https:// about. fb. com/ compa nyinfo/ Facebook. (2021b). COVID-19 and vaccine policy updates & protections. Retrieved August 11, 2022, from https:// www. faceb ook. com/ help/ 23076 48814 94641 Facebook. (2021c). Environmental, social and governance FAQs. Retrieved August 11, 2022, from https:// inves tor. fb. com/ esgresou rces/ frequ entlyaskedquest ionsesg/ defau lt. aspx Facebook. (2021d). Facebook brand resource center. Retrieved July 7, 2021, from https:// en. faceb ookbr and. com/ Facebook. (2021e). How can I use Facebook to stay updated about the coronavirus (COVID-19)? Retrieved August 11, 2022, from https:// www. faceb ook. com/ help/ 23141 63347 48066/? helpr ef= searc h& query= COVID19& search_ sessi on_ id= 8d226 8c131 b66b9 478d8 8eebd 06d64 8a& sr=2 1111 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… Facebook. (2021f). Our approach to misinformation. Retrieved May 11, 2022, from https:// trans paren cy. fb. com/ featu res/ appro achtomisin forma tion/ Facebook. (2021g). Promoting safety and expression. Retrieved August 12, 2022, from https:// about. faceb ook. com/ actio ns/ promo tingsafetyandexpre ssion/ Facebook. (2021h). Timeline: Action against COVID-19. Retrieved August 12, 2022, from https:// about. faceb ook. com/ actio ns/ respo ndingtoCOVID19/ Fukuyama, F., & Grotto, A. (2020). Comparative media regulation in the United States and Europe. In N. Persily & J. A. Tucker (Eds.), Social Media and Democracy: The State of the Field, Prospects for Reform (pp. 199–219). Cambridge University Press. https:// doi. org/ 10. 1017/ 97811 08890 960 Geeng, C., Yee, S., & Roesner, F. (2020). Fake news on Facebook and Twitter: Investigating how people (don’t) investigate. In R. Bernhaupt, F. Mueller, D. Verweij, & J. Andres (Eds.), Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems (pp. 1–14). New York, NY, USA: ACM. https:// doi. org/ 10. 1145/ 33138 31. 33767 84 Ghosh, D. (2021). Are we entering a new era of social media regulation? Harvard Business Review. Retrieved August 11, 2022, from https:// hbr. org/ 2021/ 01/ areweenter ing-aneweraofsocialmediaregul ation Gioia, D. A., Corley, K. G., & Hamilton, A. L. (2013). Seeking qualitative rigor in inductive research. Organizational Research Methods, 16(1), 15–31. https:// doi. org/ 10. 1177/ 10944 28112 452151 Gooch, A. (2020). Fighting disinformation: A key pillar of the COVID-19 recovery. OECD Forum. Retrieved May 8, 2022, from http:// www. oecdforum. org/ posts/ fight ingdisin forma tion-akeypillaroftheCOVID19recov ery Google. (2006). Google to acquire YouTube for $1.65 billion in stock. Retrieved May 8, 2022, from http:// googl epress. blogs pot. com/ 2006/ 10/ googletoacqui reyoutu befor165_ 09. html GRI. (2014). GRI G4 media sector disclosures. Retrieved May 8, 2022, from https:// www. globa lrepo rting. org/ searc h/? query= G4+ Media+ Sector Gunningham, N., & Rees, J. (1997). Industry self-regulation: An institutional perspective. Law & Policy, 19(4), 363–414. https:// doi. org/ 10. 1111/ 14679930. t01-100033 Héroux, S., & Fortin, A. (2014). Exploring IT dependence and IT governance. Information Systems Management, 31(2), 143–166. https:// doi. org/ 10. 1080/ 10580 530. 2014. 890440 Horn,N. (2021). Grundlagen der digitalen Ethik – eine normative Orientierung in der vernetzten Welt. Stiftung Datenschutz. Leipzig. Retrieved August 10, 2022, from https:// stift ungda tensc hutz. org/ filea dmin/ Redak tion/ Dokum ente/ Digit ale_ Ethik/ SDS_ Brosc huere_ Digit ale_ Ethik_ Downl oad. pdf Huber, T. L., Kude, T., & Dibbern, J. (2017). Governance practices in platform ecosystems: Navigating tensions between cocreated value and governance costs. Information Systems Research, 28(3), 563–584. https:// doi. org/ 10. 1287/ isre. 2017. 0701 Jin, K. X. (2020). Keeping People Safe and Informed About the Coronavirus. Facebook. Retrieved July 14, 2021, from https:// about. fb. com/ news/ 2020/ 12/ coron avirus/ Jørgensen, R. F., & Zuleta, L. (2020). Private governance of freedom of expression on social media platforms: EU content regulation through the lens of human rights standards. Nordicom Review, 41(1), 51–67. https:// doi. org/ 10. 2478/ nor20200003 Kompetenzzentrum Öffentliche IT (2016). Digitalisierung des Öffentlichen. Berlin. Retrieved May 11, 2022, from https:// www. oeffe ntlic heit. de/ docum ents/ 10181/ 14412/ Digit alisi erung+ des+% C3% 96ffe ntlic hen Kourula, A., Moon, J., Salles-Djelic, M. L., & Wickert, C. (2019). New roles of government in the governance of business conduct: Implications for management and organizational research. Organization Studies, 40(8), 1101–1123. https:// doi. org/ 10. 1177/ 01708 40619 852142 Lahat, L., & Sher-Hadar, N. (2020). A threefold perspective: Conditions for collaborative governance. Journal of Management and Governance, 24(1), 117–134. https:// doi. org/ 10. 1007/ s1099701909465-1 Lazer, D. M. J., Baum, M. A., Benkler, Y., Berinsky, A. J., Greenhill, K. M., Menczer, F., Metzger, M. J., Nyhan, B., Pennycook, G., Rothschild, D., Schudson, M., Sloman, S. A., Sunstein, C. R., Thorson, E. A., Watts, D. J., & Zittrain, J. L. (2018). The science of fake news. Science, 359(6380), 1094– 1096. https:// doi. org/ 10. 1126/ scien ce. aao29 98 Lindman, J., Makinen, J., & Kasanen, E. (2023). Big Tech’s power, political corporate social responsibility and regulation. Journal of Information Technology, 38(2), 144–159. https:// doi. org/ 10. 1177/ 02683 96222 11135 96 1112 L.Warnke et al. 1 3 Lovell, T. (2020). NHS joins forces with tech firms to stop the spread of COVID-19 misinformation. Healthcare IT News. Retrieved August 10, 2022, from https:// www. healt hcare itnews. com/ news/ emea/ nhsjoinsforcestechfirmsstopspreadcovid19misin forma tion Lütjen, T. (2016). Die Politik der Echokammer: Wisconsin und die ideologische Polarisierung der USA. Transcript Verlag. https:// doi. org/ 10. 14361/ 97838 39436 073 Marin, L. (2020). Three contextual dimensions of information on social media: Lessons learned from the COVID-19 infodemic. Ethics and Information Technology. https:// doi. org/ 10. 1007/ s1067602009550-2 Marin, L. (2021). Sharing (mis) information on social networking sites. An exploration of the norms for distributing content authored by others. Ethics and Information Technology, 23(3), 363–372. https:// doi. org/ 10. 1007/ s1067602109578-y Meta. (2022). Form 10-K: Annual report pursuant to section13 or 15(d) of the securities exchange act of 1934 for the fiscal year ended December 31, 2021. Retrieved September 27, 2022, from https:// d18r n 0p25n wr6d. cloud front. net/ CIK00013 26801/ 14039 b472e2f40549dc571bcc 7cf01 ce. pdf Monaghan, S., Tippmann, E., & Coviello, N. (2020). Born digitals: Thoughts on their internationalization and a research agenda. Journal of International Business Studies, 51, 11–22. https:// doi. org/ 10. 1057/ s4126701900290-0 Mosseri, A. (2017). Working to stop misinformation and false news. Meta. Retrieved September 27, 2022, from https:// www. faceb ook. com/ forme dia/ blog/ worki ngtostopmisin forma tionandfalsenews Muric, G., Wu, Y., & Ferrara, E. (2021). COVID-19 vaccine hesitancy on social media: Building a public Twitter data set of antivaccine content, vaccine misinformation, and conspiracies. JMIR Public Health and Surveillance, 7(11), e30642. https:// doi. org/ 10. 2196/ 30642 Nolan, R., & McFarlan, F. W. (2005). Information technology and the boards of directors. Harvard Business Review, 83(10), 96–106. Oversight Board. (2021). Case decision 2020–006-FB-FBR. Retrieved May 11, 2022, from https:// overs ightb oard. com/ decis ion/ FBXWJQB U9A/ Pariser, E. (2011). The filter bubble: What the internet is hiding from you. Penguin Press. Parker, G. G., van Alstyne, M. W., & Choudary, S. P. (2016). Platform revolution: How networked markets are transforming the economy and how to make them work for you. W. W. Norton & Company. Pennycook, G., Bear, A., Collins, E. T., & Rand, D. G. (2020a). The implied truth effect: Attaching warnings to a subset of fake news headlines increases perceived accuracy of headlines without warnings. Management Science, 66(11), 4921–5484. https:// doi. org/ 10. 1287/ mnsc. 2019. 3478 Pennycook, G., McPhetres, J., Zhang, Y., Lu, J. G., & Rand, D. G. (2020b). Fighting COVID-19 misinformation on social media: Experimental evidence for a scalable accuracy-nudge intervention. Psychological Science, 31(7), 770–780. https:// doi. org/ 10. 1177/ 09567 97620 939054 Reuters Institute for the Study of Journalism. (2021a). Digital news report—interactive—source of news: Social media. Retrieved May 3, 2021, from https:// www. digit alnew srepo rt. org/ inter active/. Reuters Institute for the Study of Journalism. (2021b) Global active usage penetration of leading social networks as of February 2021. InStatista. Retrieved May 11, 2022, from https:// www. stati sta. com/ stati stics/ 274773/ globalpenet rationofselec tedsocialmediasites/ Ridder, H.-G. (2020). Case study research: Approaches, methods, contribution to theory (2nd ed.). Rainer Hampp Verlag. Roberts, J. J. (2020). Facebook’s new tool to stop fake news is a game changer—if the company would only use it. Fortune. Retrieved September 19, 2022, from https:// fortu ne. com/ 2020/ 10/ 18/ faceb ooktoolstopfakenewsviralconte ntreviewsystemfbbusin essmodel/ Roozenbeek, J., Schneider, C. R., Dryhurst, S., Kerr, J., Freeman, A. L. J., Recchia, G., van der Bles, A. M., & van der Linden, S. (2020). Susceptibility to misinformation about COVID-19 around the world. Royal Society Open Science, 7(10), 201199. https:// doi. org/ 10. 1098/ rsos. 201199 Rosen, G. (2020). An update on our work to keep people informed and limit misinformation about COVID-19. Meta. Retrieved September 10, 2021, from https:// about. fb. com/ news/ 2020/ 04/ COVID19misin foupdate/ Rosen, G. (2021). How we’re tackling misinformation across our apps. Meta. Retrieved September 10, 2021, from https:// about. fb. com/ news/ 2021/ 03/ howweretackl ingmisin forma tionacrossourapps/ Royakkers, L., Timmer, J., Kool, L., & van Est, R. (2018). Societal and ethical issues of digitization. Ethics and Information Technology, 20(2), 127–142. https:// doi. org/ 10. 1007/ s106760189452-x 1113 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… SASB. (2018). Internet media & services—Sustainability accounting standard. Retrieved June 17, 2021, from https:// www. sasb. org/ wpconte nt/ uploa ds/ 2018/ 11/ Inter net_ Media_ Servi ces_ Stand ard_ 2018. pdf Schrempf-Stirling, J., & Wettstein, F. (2023). The mutual reinforcement of hard and soft regulation. Academy of Management Perspectives, 37(1), 72–90. https:// doi. org/ 10. 5465/ amp. 2022. 0029 Srnicek, N. (2017). The challenges of platform capitalism: Understanding the logic of a new business model. Juncture, 23(4), 254–257. https:// doi. org/ 10. 1111/ newe. 12023 Steinert, S. (2020). Corona and value change. The role of social media and emotional contagion. Ethics and Information Technology. https:// doi. org/ 10. 1007/ s1067602009545-z Törnberg, P. (2018). Echo chambers and viral misinformation: Modeling fake news as complex contagion. PLoS ONE, 13(9), e0203958. https:// doi. org/ 10. 1371/ journ al. pone. 02039 58 Trittin-Ulbrich, H., Scherer, A. G., Munro, I., & Whelan, G. (2021). Exploring the dark and unexpected sides of digitalization: Toward a critical agenda. Organization, 28(1), 8–25. https:// doi. org/ 10. 1177/ 13505 08420 968184 Turillazzi, A., Taddeo, M., Floridi, L., & Casolari, F. (2023). The digital services act: An analysis of its ethical, legal, and social implications. Law, Innovation and Technology, 15(1), 83–106. https:// doi. org/ 10. 1080/ 17579 961. 2023. 21841 36 Twitter. (2021a). Coronavirus: Staying safe and informed on Twitter. Retrieved May 11, 2022, from https:// blog. twitt er. com/ en_ us/ topics/ compa ny/ 2020/ covid19 Twitter. (2021b). COVID-19 misleading information policy. Retrieved May 11, 2022, from https:// help. twitt er. com/ en/ rulesandpolic ies/ medic almisin forma tionpolicy Twitter. (2021c). Glossary. Retrieved May 12, 2022, from https:// help. twitt er. com/ en/ resou rces/ gloss ary Twitter. (2021d). Twitter’s services, corporate affiliates, and your privacy. Retrieved May 12, 2022, from https:// help. twitt er. com/ en/ rulesandpolic ies/ twitt erservi cesandcorpo rateaffil iates Twitter. (2022). Form 10-K: annual report pursuant to section13 or 15(d) of the securities exchange act of 1934 for the fiscal year ended December 31, 2021. Retrieved September 27, 2022, from https:// s22. q4cdn. com/ 82664 1620/ files/ doc_ finan cials/ 2021/ ar/ Fisca lYR20 21_ Twitt er_ Annua l_Report. pdf on Twitter Safety. (2020). COVID-19: Our approach to misleading vaccine information. Retrieved September 27, 2022, from https:// blog. twitt er. com/ en_ us/ topics/ compa ny/ 2020/ covid 19vacci ne Twitter Safety. (2021). Updates to our work on COVID-19 vaccine misinformation. Retrieved May 11, 2022, from https:// blog. twitt er. com/ en_ us/ topics/ compa ny/ 2021/ updat estoourworkonCOVID19vacci nemisin forma tion van Dijck, J. (2013). The culture of connectivity: A critical history of social media. Oxford University Press. van Dijck, J., Poell, T., & de Waal, M. (2018). The platform society: Public values in a connective world. Oxford University Press. Van Grembergen, W., & De Haes, S. (2009). Enterprise governance of information technology. Springer. https:// doi. org/ 10. 1007/ 978-038784882-2 Van Maanen, J., Sørensen, J. B., & Mitchell, T. R. (2007). The interplay between theory and method. Academy of Management Review, 32(4), 1145–1154. https:// doi. org/ 10. 5465/ amr. 2007. 26586 080 Vosoughi, S., Roy, D., & Aral, S. (2018). The spread of true and false news online. Science, 359(6380), 1146–1151. https:// doi. org/ 10. 1126/ scien ce. aap95 59 We Are Social, DataReportal & Meltwater. (2023). Most popular social networks worldwide as of January 2023, ranked by number of monthly active users (in millions). In Statista. Retrieved June 20, 2023, from https:// www. stati sta. com/ stati stics/ 272014/ globalsocialnetwo rksrankedbynumberofusers/ Wilkin, C. L., & Chenhall, R. H. (2020). Information technology governance: Reflections on the past and future directions. Journal of Information Systems, 34(2), 257–292. https:// doi. org/ 10. 2308/ isys52632 World Health Organization. (2020). COVID-19 Mythbusters. Retrieved August 11, 2022, from https:// www. who. int/ emerg encies/ disea ses/ novelcoron avirus2019/ adviceforpublic/ mythbuste rs World Health Organization. (2021a). Coronavirus disease (COVID-19) - Q&A. Retrieved May 11, 2022, from https:// www. who. int/ newsroom/qadetail/ coron avirusdisea seCOVID19 World Health Organization. (2021b). Let’s flatten the infodemic curve. Retrieved May 4, 2023, from https:// www. who. int/ newsroom/ spotl ight/ let-sflatt entheinfod emiccurve World Health Organization. (2022). WHO Coronavirus (COVID-19) dashboard. Retrieved September 11, 2022, from https:// covid 19. who. int 1114 L.Warnke et al. 1 3 YouTube. (2019). The four Rs of responsibility, part 2: Raising authoritative content and reducing borderline content and harmful misinformation. YouTube Official Blog. Retrieved September 10, 2021, from https:// blog. youtu be/ insideyoutu be/ thefourrsofrespo nsibi lityraiseandreduce/ YouTube. (2021a). COVID-19 medical misinformation policy. YouTube Help. Retrieved April 11, 2022, from https:// suppo rt. google. com/ youtu be/ answer/ 98917 85? hl= de& hl= en& ref_ topic= 92824 36 YouTube. (2021b). How does YouTube combat misinformation?—borderline content. Retrieved April 5, 2022, from https:// www. youtu be. com/ intl/ en_ us/ howyo utube works/ ourcommi tments/ fight ingmisin forma tion/# borde rlineconte nt YouTube. (2021c). How does YouTube combat misinformation?—determining misinfo. Retrieved April 5, 2022, from https:// www. youtu be. com/ intl/ en_ us/ howyo utube works/ ourcommi tments/ fight ingmisin forma tion/# deter miningmisin fo YouTube. (2021d). How does YouTube combat misinformation?—raising quality info. Retrieved April 5, 2022, from https:// www. youtu be. com/ intl/ en_ us/ howyo utube works/ ourcommi tments/ fight ingmisin forma tion/# raisi ngquali tyinfo YouTube. (2021e). Progress on managing harmful content—Detection Source. Retrieved April 5, 2022, from https:// www. youtu be. com/ intl/ en_ us/ howyo utube works/ progr essimpact/ respo nsibi lity/# detec tionsource YouTube. (2021f). Progress on managing harmful content—Removal by views. Retrieved April 5, 2022, from https:// www. youtu be. com/ intl/ en_ us/ howyo utube works/ progr essimpact/ respo nsibi lity/# remov albyviews YouTube. (2021g). Unsere Fortschritte beim Umgang mit schädlichen Inhalten—Entfernte Videos nach Aufrufen. Retrieved April 5, 2022, from https:// www. youtu be. com/ intl/ ALL_ de/ howyo utube works/ progr essimpact/ respo nsibi lity/# remov albyviews Zarsky, T. (2016). The trouble with algorithmic decisions: an analytic road map to examine efficiency and fairness in automated and opaque decision making. Science, Technology, & Human Values, 41(1), 118–132. https:// doi. org/ 10. 1177/ 01622 43915 605575 Zuboff, S. (2015). Big other: surveillance capitalism and the prospects of an information civilization. Journal of Information Technology, 30(1), 75–89. https:// doi. org/ 10. 1057/ jit. 2015.5 Zuboff, S. (2019). The age of surveillance capitalism: The fight for a human future at the new frontier of power. Public Affairs. Zuboff, S. (2022). Surveillance capitalism or democracy? The death match of institutional orders and the politics of knowledge in our information civilization. Organization Theory, 3(3), 1–79. https:// doi. org/ 10. 1177/ 26317 87722 11292 90 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Lina Warnke is a research associate at the Helmut Schmidt University Hamburg (Germany). Her research interests focus on sustainability reporting, corporate governance and corporate social responsibility. Anna‑Lena Maier is a project and change manager at the ma-co maritimes competenzcentrum GmbH and oversees transformation projects related to digitalization and the energy transition. She also is a lecturer and researcher covering topics such as business and government, business and peace, political CSR and responsible management. She received her PhD from the University of Hamburg (Germany), where she taught several courses on (international) strategic management, business ethics and CSR. Dirk Ulrich Gilbert is a professor of business ethics and management at the University of Hamburg (Germany). He received his PhD from the University of Frankfurt and held positions at the University of New South Wales and the University of Nuremberg. His most recent research focuses on international accountability standards, labour rights in global supply chains, political CSR, and responsible management education. He has published in internationally acclaimed journals, such as Business Ethics Quarterly, Business and Society, Academy of Management Learning and Education, Journal of Management Inquiry, Management International Review, and the Journal of Business Ethics. 1115 1 3 Social media platforms’ responses toCOVID‑19‑related mis‑… Authors and Affiliations LinaWarnke1 · Anna‑LenaMaier2 · DirkUlrichGilbert3 * Lina Warnke lina.war[email protected] Anna-Lena Maier [email protected] Dirk Ulrich Gilbert [email protected] 1 Faculty ofEconomics andSocial Sciences, Helmut-Schmidt-Universität / Universität Der Bundeswehr Hamburg, Holstenhofweg 85, 22043Hamburg, Germany 2 Ma-Co Maritimes Competenzcentrum, Köhlbranddeich 30, 20457Hamburg, Germany 3 Department ofSocioeconomics, Faculty ofBusiness, Economics andSocial Sciences, Universität Hamburg, Von-Melle-Park 9, 20146Hamburg, Germany