PurposePurveyors of fake news perpetuate information that can harm society, including businesses. Social media's reach quickly amplifies distortions of fake news. Research has not yet fully explored the mechanisms of such adversarial behavior or the adversarial techniques of machine learning that might be deployed to detect fake news. Debiasing techniques are also explored to combat against the generation of fake news using adversarial data. The purpose of this paper is to present the challenges and opportunities in fake news detection.Design/methodology/approachFirst, this paper provides an overview of adversarial behaviors and current machine learning techniques. Next, it describes the use of long short-term memory (LSTM) to identify fake news in a corpus of articles. Finally, it presents the novel adversarial behavior approach to protect targeted business datasets from attacks.FindingsThis research highlights the need for a corpus of fake news that can be used to evaluate classification methods. Adversarial debiasing using IBM's Artificial Intelligence Fairness 360 (AIF360) toolkit can improve the disparate impact of unfavorable characteristics of a dataset. Debiasing also demonstrates significant potential to reduce fake news generation based on the inherent bias in the data. These findings provide avenues for further research on adversarial collaboration and robust information systems.Originality/valueAdversarial debiasing of datasets demonstrates that by reducing bias related to protected attributes, such as sex, race and age, businesses can reduce the potential of exploitation to generate fake news through adversarial data.

PurposeThe purpose of this study is to investigate the impact of the COVID-19 pandemic on the beliefs and attitudes toward the use of information and communication technology (ICT). The study examines the challenges of implementing ICT-based training and provides insights for promoting the acceptance of online training in volunteer sports communities.Design/methodology/approachThe study uses an action design research methodology that combines the implementation of ICT-based training, interviews, and a survey of 523 participants to examine the influence of online training on beliefs and attitudes.FindingsThe study shows that before the COVID-19 pandemic, soccer referees had negative beliefs about the use of ICT for learning. However, the experience of being forced to use ICT for training during the pandemic led to a positive shift in their beliefs about ICT.Research limitations/implicationsThe study offers four lessons learned for promoting the use of ICT-based training in voluntary sports. Future research should investigate the influence of blended learning approaches on affective, cognitive, and skill-based learning outcomes.Practical implicationsThe study has practical implications for those responsible for implementing ICT-based training in voluntary sport. The findings suggest that design features such as usefulness, ease of use and enjoyment should be emphasized to increase the acceptance of online training.Originality/valueThe study contributes to the literature by providing insights into the challenges of implementing ICT-based training in voluntary sport contexts. The findings suggest that the experience of being forced to use ICT can promote the acceptance of online training in volunteer sports communities.

PurposeWhile the metaverse is promised to be the next big step for the Internet, this new technology may also bear negative impacts on individuals and society. Drawing on innovation resistance literature, this article explores the reasons for metaverse resistance.Design/methodology/approachThe study is based on 66 semi-structured interviews, and the subsequent data were analysed thematically.FindingsThe findings revealed 11 reasons for metaverse resistance: lack of understanding, lack of regulation, addiction avoidance, claustrophobia, loss of social ties, disconnection from reality, privacy concerns, extreme consumer society, unseen benefits, infeasibility and nausea.Practical implicationsBy understanding the various reasons for metaverse resistance managers and policymakers can make better decisions to overcome the challenges facing this innovation, rather than adopting a “one-size-fits-all” approach.Originality/valueWhile the literature has mainly adopted a positive perspective on the metaverse, this research offers a more nuanced view by identifying the reasons why consumers may resist the metaverse. Furthermore, this study introduces for the first-time “addiction-driven-innovation-resistance (ADIR)” as a potential reason for metaverse resistance, which may also apply to other cases of innovation resistance, when new innovations are perceived as being “too good” and therefore potentially addictive.

PurposeMisinformation is notoriously difficult to combat. Although social media firms have focused on combating the publication of misinformation, misinformation accusations, an important by-product of the spread of misinformation, have been neglected. The authors offer insights into factors contributing to the spread of misinformation accusations on social media platforms.Design/methodology/approachThe authors use a corpus of 234,556 tweets about the 2020 US presidential election (Study 1) and 99,032 tweets about the 2022 US midterm elections (Study 2) to show how the sharing of misinformation accusations is explained by locomotion orientation.FindingsThe study findings indicate that the sharing of misinformation accusations is explained by writers' lower locomotion orientation, which is amplified among liberal tweet writers.Research limitations/implicationsPractitioners and policymakers can use the study findings to track and reduce the spread of misinformation accusations by developing algorithms to analyze the language of posts. A limitation of this research is that it focuses on political misinformation accusations. Future research in different contexts, such as vaccines, would be pertinent.Practical implicationsThe authors show how social media firms can identify messages containing misinformation accusations with the potential to become viral by considering the tweet writer's locomotion language and geographical data.Social implicationsEarly identification of messages containing misinformation accusations can help to improve the quality of the political conversation and electoral decision-making.Originality/valueStrategies used by social media platforms to identify misinformation lack scale and perform poorly, making it important for social media platforms to manage misinformation accusations in an effort to retain trust. The authors identify linguistic and geographical factors that drive misinformation accusation retweets.

PurposeThis study aims to propose a model that delineated the diffusion process of product-harm misinformation on social media. Drawing on theoretical insights from cue diagnosticity and corporate associations, the proposed model mapped out how consumers' information skepticism and perceived content credibility influence their perceived diagnosticity of the product-harm misinformation and corporate ability (CA) associations with the company being impacted, which in turn influenced their trust toward the company and negative word-of-mouth (NWOM) intention.Design/methodology/approachA survey was conducted with 504 US consumers to empirically test the proposed model. Following the survey, in-depth interviews were conducted with 11 communication professionals regarding the applicability of the model.FindingsWhen exposed to product-harm misinformation on social media, consumers' perceived diagnosticity of misinformation was negatively impacted by their information skepticism and positively impacted by perceived content credibility of misinformation. Perceived diagnosticity of product-harm misinformation negatively impacted consumers' CA associations, which then led to decreased trust and increased NWOM intention. Findings from the interviews further supported the diffusion process and provided insights on strategies to combat product-harm misinformation. Strategies shared by the interviewees included preparedness and social listening, proactive outreach and building strong CA associations as preventative measures.Originality/valueThis study incorporates the theoretical frameworks of cue diagnosticity and corporate associations into the scholarship of misinformation and specifically addresses the unique diffusion process of product-harm misinformation on social media. This study provides insights and tangible recommendations for communication professionals to combat product-harm misinformation.

PurposeSocial commerce (s-commerce) offers community-based platforms that facilitate customer-to-customer interactions and the development of customers' social shopping-based experience. While prior research has addressed the role of customer engagement (CE) in boosting s-commerce-based sales and performance, insight into the effect of s-commerce attributes on CE remains tenuous. Addressing this gap, this study examines the role of specific s-commerce attributes (i.e. community, collaboration, interactivity and social dynamics) on CE, which is, in turn, proposed to impact customers' repurchase- and electronic word of mouth (eWOM) intention.Design/methodology/approachA web-based survey was deployed to target users of a popular s-commerce platform, Etsy.com. Partial least squares structural equation modeling (PLS-SEM) was, then, used to analyze the survey data collected from 390 users.FindingsThe results reveal that the four examined attributes positively affect CE. The findings also demonstrate CE's positive effect on customers' repurchase- and eWOM intention.Originality/valueThough CE has been identified as a key s-commerce performance indicator, little remains known about the role of specific s-commerce attributes in driving CE, as, therefore, explored in this research. Specifically, the authors examine the role of s-commerce-based community, collaboration, interactivity and social dynamics on CE. Their analyses also corroborate that CE, in turn, drives customers' post-purchase (i.e. repurchase/eWOM) intention. Managerially, our findings can be used to develop more engaging s-commerce platforms.

PurposeOne of the most important challenges confronting enterprise managers is that of controlling employees' social cyberloafing. The use of enterprise social media entails opportunities for cyberloafing. However, previous research on how enterprise social media use affects cyberloafing is rather limited. Using the job demands-resources (JD-R) model, this paper proposes a research model to investigate the relationship between enterprise social media usage and employees' social cyberloafing behavior.Design/methodology/approachStructural equation modeling was performed to test the research model and hypotheses. Surveys were conducted in an online platform in China, generating 510 employees' data for analysis.FindingsFirst, both public social media and private social media used for work-related and social-related purposes have a positive effect on employees' job engagement. Further, job engagement has a negative effect on employees' social cyberloafing. Second, the use of public social media for work-related and social-related purposes has no effect on employees' emotional exhaustion. However, work-related private social media usage has a negative effect on employees' emotional exhaustion, and social-related private social media usage has a positive effect on employees' emotional exhaustion. Further, employees' emotional exhaustion has a positive effect on employees' social cyberloafing. Third, there are significant differences in the effects of enterprise social media on employees' social cyberloafing between male and female employees.Originality/valueFirst, this paper contributes to the social cyberloafing literature by establishing a relationship between enterprise social media usage and social cyberloafing in relation to the dual influence mechanism. Second, it contributes to the JD-R model by clarifying how the use of enterprise social media with different motivations affects social cyberloafing through a mediation mechanism, namely, an enabling mechanism and a burden mechanism. Third, this paper also contributes to the social cyberloafing literature by revealing the boundary condition, namely gender, between enterprise social media use and employees' social cyberloafing.