The Sound of Progress: Investigating the Representation of Artificial Intelligence in Music

2023 | International Conference on Business Informatics | Citations: 0

Authors: Banh, Leonardo; Strobel, Gero

Abstract: The role of artificial intelligence in daily life is constantly advancing and h ... Expand

Abstract: The role of artificial intelligence in daily life is constantly advancing and has become an important topic of discussion. With music and its lyrics being a vehicle to express topics of society, this paper investigates how artificial intelligence is perceived by musicians and reflected in their songs. By analyzing the lyrics of over 1,200 songs over three decades, this work applies sentiment analysis to extract polarities and emotions. The results provide insights into how musicians view and reflect the impact of artificial intelligence on society and how this is reflected in their song texts. The findings show an increase in songs mentioning artificial intelligence-related terms, with a trend of more songs implying negativity, such as anger. However, minor increases in positive emotions indicate musicians' ambivalent views on hopes and fear of artificial intelligence. Collapse

Topics: artificial intelligence deepfake database system

Methods: sentiment analysis deep learning machine learning language model simple linear regression

Theories: communication theory

Navigating Uncertain Waters: How Organizations Respond to Institutional Pressure in Times of the Looming EU AI Act

2023 | International Conference on Information Systems | Citations: 0

Authors: Kahdan, Manoj; Hartwich, Nicole; Salge, Oliver; Cichy, Patrick

Abstract: The looming AI Act puts significant pressure on organizations to develop safe a ... Expand

Abstract: The looming AI Act puts significant pressure on organizations to develop safe and reliable AI. First, we analyze 975 AI incidents to provide an overview of the actors involved in identifying primary AI failures. Second, using institutional theory and based on over two years of observation of more than 400 actors, we identify four strategies (circumvent, comply, compromise, and control) and exemplary response practices (relabel, redefine, balance, and predefine) for how firms respond to institutional pressures from AI regulation. We thereby extend the institutional theory perspective on AI regulation. Using our longitudinal multi-actor perspective, we develop the 4C-Framework for how organizations actively respond to AI regulation. Moreover, we demonstrate how this framework can be applied to impending AI regulation, including strategies for engaging with regulators, exploring opportunities for collaboration, and dealing with the spread of misinformation. Overall, this paper provides a comprehensive approach for companies to proactively shape AI regulation. Collapse

Topics: artificial intelligence database system misinformation innovation management responsible artificial intelligence

Methods: longitudinal research case study qualitative interview qualitative content analysis cluster analysis

Theories: institutional theory

Deepfake Valence: A Double-edged Sword

2023 | Americas Conference on Information Systems | Citations: 0

Authors: Liang, Huigang; Chew, Shao

Abstract: Deepfakes have garnered widespread attention for their malicious use. Despite i ... Expand

Abstract: Deepfakes have garnered widespread attention for their malicious use. Despite its infamy, deepfakes can bring forth a myriad of benefits. Given the controversies around deepfake, it is unclear why people adopt this technology for various purposes. Through the lens of rational choice theory, we develop a research model to explain deepfake adoption. From the perspective of deepfake developers, we categorize deepfake usage based on two dimensions: purpose (utilitarian vs. hedonic) and valence (benevolent vs. malevolent). Deepfake adoption is driven by a risk benefit analysis which, in turn, is influenced by socioenvironmental factors. Collapse

Topics: deepfake social norm technology adoption computer crime

Methods: SEM tool survey statistical hypothesis test scale reliability field study

Theories: rational choice theory big five model

AI-Generated Voice in Short Videos: A Digital Consumer Engagement Perspective

2023 | International Conference on Information Systems | Citations: 0

Authors: Luo, Jihao; Zheng, Chenxu; Yin, Jiamin; Teo, Hock-Hai

Abstract: The AI-generated voice (AIGV) has been widely applied in the short video indust ... Expand

Abstract: The AI-generated voice (AIGV) has been widely applied in the short video industry to facilitate video creation. However, the impact of AIGV on digital consumer engagement (DCE) remains unclear. In light of this, the study investigates the effect of AIGV on DCE by analyzing a panel dataset with observations of 21,541 videos for 3,647 content creators on TikTok. Preliminary results of a series of fixed-effect panel regressions reveal that using AIGV has a significantly negative effect on DCE (5.4% reduction in the number of likes, 5.2% reduction in the number of comments, and 7.4% reduction in the number of shares). Our further analyses show that this negative effect is particularly significant at the rising action stage of short videos. With these findings, this study is expected to have theoretical contributions to the literature on short videos and practical implications about the appropriate usage of AIGV. Collapse

Topics: TikTok artificial intelligence deepfake application programming interface YouTube

Methods: computational algorithm machine learning propensity score matching experimental group theoretical contribution

Theories: dual-system theory

Battling Disinformation Intermediaries: An Analysis of Information Policies

2023 | Americas Conference on Information Systems | Citations: 0

Authors: Zhu, Xiaohua; Yang, Shengnan

Abstract: Despite the increasing scholarly attention to disinformation, there is a lack o ... Expand

Abstract: Despite the increasing scholarly attention to disinformation, there is a lack of research on policies of disinformation, which has extended beyond social media to cover other aspects of disinformation distribution. This study is an investigation into a specific aspect of disinformation policy, intermediaries, defined as all channels, carriers, means, and mechanisms that enable disinformation creation, distribution, transmission, dissemination, and consumption, using technology or through traditional methods. Using an inductive thematic analysis approach, this study examined existing information policies in China and the United States and identified a wide range of disinformation intermediaries that are currently regulated or have the potential to be regulated by the government. The paper reports on these intermediaries and offers an analysis of the contexts of disinformation regulation in the two countries. The findings provide empirical evidence on current disinformation policy practices that may assist researchers, policymakers, and legislators in understanding the current disinformation policy landscape. Collapse

Topics: disinformation misinformation social media IT policy database system

Methods: qualitative content analysis qualitative coding archival research card sorting

Misinformation: A Survey of State-of-Art and Future Research Opportunities

2022 | Americas Conference on Information Systems | Citations: 0

Authors: Jiang, Wenting; Gupta, Ashish; Taylor, Gabrielle; Qin, Xiao

Abstract: The proliferation of misinformation has known detrimental impacts on individual ... Expand

Abstract: The proliferation of misinformation has known detrimental impacts on individuals, organizations, and society. The use of social media further aggravates misinformation spread across broad sectors of populations. The quality of misinformation has also improved over recent years, making it difficult for various Artificial Intelligence (AI) based detection techniques to differentiate the truth from the fake. There is a dearth of systematic literature reviews on this topic. This survey study conducts a comprehensive examination of recent work done on misinformation to synthesize four themes where recent research on misinformation has been evolving: (1) spread of misinformation, (2) impacts of misinformation, (3) misinformation detection, and (4) mitigation of misinformation. This effort serves to provide guidance for researchers in pursuing this important line of research on developing improved models and analytics approaches that help mitigate the impact of misinformation. Collapse

Topics: misinformation fake news disinformation deepfake open source

Methods: literature sample survey structured literature research natural language processing meta analysis

Deepfake Audio Detection

2022 | Americas Conference on Information Systems | Citations: 0

Authors: Khan, Madeeha; Goel, Sanjay; Anandan, Jaswant Katar; Zhao, Jersey; Naik, Ramavath Rakesh

Abstract: Deepfakes, algorithms that use Machine Learning (ML) to generate fake yet reali ... Expand

Abstract: Deepfakes, algorithms that use Machine Learning (ML) to generate fake yet realistic content, represent one of the premier security challenges in the 21st century. Deepfakes are not limited to just videos, as deepfake audio is a fast-growing field with an enormous number of applications. Recently, multiple Convolutional Neural Network (CNN) based techniques have been developed that generate realistic results that are difficult to distinguish from actual speech. In this work, we extracted audio features from real and synthesized audio files and determined that Mel-Frequency Cepstral Coefficients (MFCCs) in synthesized audio show a significant difference from the MFCCs in real audio. Using Deep Neural Networks (DNNs), experiments were conducted to train classifiers to detect synthesized audio in different datasets, with highly successful results. Collapse

Topics: spoofing deepfake IT security threat YouTube open source

Methods: computational algorithm deep learning machine learning recurrent neural network synthesis

Towards a Theory of Digital Stigma and Deep Fake Video Technology Stigmatization in a Digitally Mediated Environment

2022 | Americas Conference on Information Systems | Citations: 0

Authors: Mohammed, Fatima; Salam, A. F.

Abstract: As social structures and IT becomes intertwined, there is a need for new theore ... Expand

Abstract: As social structures and IT becomes intertwined, there is a need for new theoretical perspectives that advance understanding of this intertwinement and its implications for IT and social actors, institutions, and society. This research attempts to develop a theory that conceptualizes the role of technology in creating and manipulating stigma symbols and the subsequent consequences of the resultant symbolic interaction on the social actors, technology, and society. Understanding the changing nature of stigma due to technology is essential in understanding resulting societal manifestations and aids in the timely management of societal implications of the dark side of technology. Collapse

Topics: deepfake social norm social identity information system use

Methods: theory development computational algorithm archival research

Theories: symbolic interaction theory

Deepfakes: An Integrative Review of the Literature and an Agenda for Future Research

2022 | Communications of the Association for Information Systems | Citations: 0

Authors: Vasist, Pramukh Nanjundaswamy; Krishnan, Satish

Abstract: We are witnessing a growing concern around the impact of hyper-realistic synthe ... Expand

Abstract: We are witnessing a growing concern around the impact of hyper-realistic synthetic media and its dissemination in what is widely known as "deepfakes." However, the phenomenon's relative newness and the fragmented nature of existing research on it across several disciplines leave a lot to be desired. In addition, empirical research is scarce, and deepfake literature is moving forward in a variety of directions without a strong theoretical foundation. The fragmented nature of extant literature on the phenomenon calls for consolidation in order to produce a thorough and current summary of deepfake research to date and map its present intellectual boundaries. We offer an integrative overview of the existing corpus of research on deepfakes in this paper, noting the wide range of domains, samples, and approaches used. We point out various gaps in deepfake narratives, including definitional concerns, a lack of comprehensive demographic and cross-geographic coverage, a lack of theoretical underpinning, thematic tensions, and imbalances in the extant literature on deepfakes. In the last section of the paper, we propose future research directions, which include a set of themes and research questions and a theoretical framework to guide future research on the topic. Deepfakes: An Integrative Review of the Literature and an Agenda for Future Research Volume 51 10.17705/1CAIS.05126 Paper 26 this domain, and practitioners may not be able to incorporate the accumulated knowledge in dealing with deepfakes. Hence, in line with our research purpose, the research questions (RQs) of our study are: RQ1: What is the status of the research profile on existing literature on deepfakes? RQ2: What are the research gaps, limitations, and recommendations for scholars and practitioners in the context of future research on deepfakes?To answer these research questions, this integrative review on deepfakes aims to review all empirical and theoretical research on deepfakes through an analysis of articles published so far on the topic in the past two decades since 2001. To identify the scholarly contribution of deepfakes to literature, we conducted a literature search of articles across leading databases for peer-reviewed literature. Although deepfakes as a term traces its origin to 2017 (Albahar & Almalki, 2019), we ensured that the literature survey accounted for publication lags, if any, and aimed to cover all articles since 2001. In the first part of this review, as a response to RQ1, we identify, synthesize, and present a current profile of extant literature on deepfakes, including annual publication trends, geographic coverage of studies and frameworks, variables, and measures related to the characterization and evolution of deepfakes. In response to RQ2, the second part of the review focuses on delineating the accumulated body of research on deepfakes since its inception to the present day and describes research findings published in literature over the past two decades since 2001. In doing so, we lay out the details of the findings from each of the selected articles on deepfakes published in peer-reviewed journals during the timeframe and highlight gaps and limitations in extant literature. We then present potential avenues for future research, as well as a state-of-the-art framework for deepfake research based on extant literature and insights gained from this integrative review.The remainder of the paper is laid out as follows: A brief background of deepfakes is presented in the second section. The third section profiles the existing literature on deepfakes and describes the methodological processes used to carry out this integrative review. Section four enlists the key aspects discussed thus far in deepfake literature, while section five reviews the empirical literature on deepfakes. This synthesis of extant literature segues into the identification of gaps and limitations in section six and forms the basis for future research themes. Section seven contains recommendations as well as a research framework for future researchers to address existing knowledge gaps. Finally, the paper concludes with a discussion of the study's theoretical and practical implications, as well as its limitations. Collapse

Topics: deepfake database system filter bubble fake news personality

Methods: literature study literature sample experiment mixed method longitudinal research

Theories: theory of reasoned action elaboration likelihood model

Levels of Digital Representation: Semiotics and the Articulation of Meaning

2021 | International Conference on Information Systems | Citations: 0

Authors: Chandra-Kruse, Leona; Seidel, Stefan; Maedche, Alexander

Abstract: Representation theory conceives of information systems (IS) as providing faithf ... Expand

Abstract: Representation theory conceives of information systems (IS) as providing faithful representations of focal real-world phenomena, such as goods, processes, or services. Some digital artifacts, however, create alternate realities, representing phenomena that are neither observable nor exist outside those artifacts (e.g., video games or virtual worlds). In such situations, established notions of IS as representations of real-world phenomena, coupled with the idea of representational fidelity to measure an IS's quality, do not apply. We use a semiotic lens and argue that in order to consider the full range of contemporary IS, including those that represent real-world phenomena and those that do not, we need to extend the focus from "IS as representations of real-world phenomena" to the idea of "IS as representations of articulated meaning." We conceptualize multiple levels of representation for real-world and digital phenomena, develop a set of propositions, and illustrate our conceptual framework using two examples. Collapse

Topics: deepfake smart wearable device mobile system

Methods: design artifact conceptual modelling multiple case study electrocardiography

Theories: semiotic theory