Digitalization and artificial knowledge for accountability in SCM: a systematic literature review

Digital transformation, through platforms, websites, social media, artificial intelligence (AI) and connected devices, has led to “datafication” (Gupta and George, 2016; Di Vaio and Varriale, 2020), which has attracted considerable attention from researchers and business practitioners. With such rapid digital progress, businesses are seeking to strengthen their decision-making, accountability and relationships with various societal actors (Ramírez and Tejada, 2019). Likewise, the COVID-19 pandemic has been an enabler of digital transformation, facilitating not only better operational performance through cost reduction and higher strategic performance but also greater opportunities to find new business markets (Wamba et al., 2022). However, the growth of the digital wave raises substantial and debatable concerns about how current industry platforms are eroding digital technologies' resilience to become data-driven and lead to transformative change (Di Vaio et al., 2020; Battisti et al., 2022). In fact, industry efforts are not enough to respond to digital waves at different levels and with different intensities (Ardito et al., 2018). Consequently, artificial knowledge has come to light as a relevant concept and choice for business operations, especially in supply chain management (SCM) (Samuel et al., 2011).

The basic idea is to use artificial knowledge to transform business models driven by innovation (Pietronudo et al., 2022) and to spread “datafication” through stakeholders (Del Giudice et al., 2023). The development and adoption of artificial knowledge is expected globally (Mikalef and Gupta, 2021), with resulting positive effects on SCM operations (Brinch et al., 2018). Some have even referred to artificial knowledge as the “life blood” of SCM since it succeeds in synchronizing knowledge information from formal and informal sources (Nayal et al., 2021). Business organizations can thus “learn by doing” via artificial knowledge (Enholm et al., 2022), the benefits of which apply to reduced product and service cycle times; in turn, this brings added value within the supply chain to offset the necessary investments for artificial knowledge adoption.

Since the pandemic, business strategy development has focused on coping with external changes by reducing risk components. Indeed, the pandemic crisis has brought attention to resilience issues (Das et al., 2022), with scholars calling for new research on resilience theory in the SCM context (Veile, 2022). AI includes the corpus of knowledge allowing machines to behave in intelligent human-like manners (Bawack et al., 2021); hence, AI can generate positive impacts on the agility, resilience and performance of a supply chain (Dubey et al., 2022). It is therefore imperative to incorporate the artificial knowledge management concept within the study of resilience in SCM (Leoni et al., 2022). However, the resilience perspective of artificial knowledge remains unclear.

Additionally, the literature has paid less attention to the link between artificial knowledge and the pillars of sustainability (Gansser and Reich, 2021). With regard to SCM, there may be a conflict between the concept of sustainability, which focuses on efficiency, and the concept of resilience, which emphasizes effectiveness (Negri et al., 2021). A potential solution is the development of a flexible value chain that holds sustainability goals as its priority (Dwivedi et al., 2021). Sustainable collaborations can be decisive in improving SCM (Le et al., 2021), given that stakeholder engagement is a vital component of sustainability reporting to achieve business legitimacy (D'Adamo, 2022). Moreover, sustainable resource management models promote competitive advantages for businesses (Appolloni et al., 2022). Hence, accountability efforts in supply chain operations are seen as a dynamic resource that assists the acceptance of the Sustainable Development Goals (SDGs) under the United Nations (UN) 2030 Agenda, which enacts environmental and social welfare policies to achieve sustainable performance (Mol, 2010). The incorporation of the accountability mechanism in SCM is interpreted as the basis of good governance that promotes organizational openness and communication (Valentinov et al., 2019), thereby enabling stakeholders to understand and take new action in supply chain operations, particularly in production and delivery processes (Gold and Heikkurinen, 2018). In contrast, without accountability, SCM operations can become vulnerable to risk, leading to poor financial performance and reporting and auditing results (Sibanda et al., 2020).

According to resilience theory, artificial knowledge examines the complex interrelationships between various digital industrial environments and operational agility (Ivanov, 2021). The theory highlights the relevant implications of artificial knowledge about SCM production and deployment, particularly those associated with novel data and information types (Sambasivan et al., 2009; Liu et al., 2013). Although AI can promote more resilient supply chains, there are shortcomings in its application (Belhadi et al., 2022). Some studies (e.g. Helo and Hao, 2022; Rana et al., 2022) have pointed out encouraging improvements, citing artificial knowledge as evidence of the evolution of digital knowledge in SCM. These developments are of great importance because of their potential impact on accountability (Kumar et al., 2020; Sibanda et al., 2020) and industrial innovation (Javaid et al., 2022). Ergo, by following sustainable principles, digitization and SCM can find mutual benefits characterized by effectiveness and efficiency (Chen et al., 2022). This underscores an evident shortcoming in the existing body of knowledge, which demands the examination of performance evaluation systems to assess both sustainability and resilience in supply chains (Shishodia et al., 2021; Hervani et al., 2022). Such systems should include basic digital transformation outcomes in assessing SCM resilience (Yin and Ran, 2022), thereby distinguishing and combining the role of technology and governance (Faruquee et al., 2021). The links between the issues introduced here are still weak from a conceptual standpoint; therefore, through the exploration and comprehension of multiple research perspectives (Saunders et al., 2015), this study sought to answer two main research questions (RQs), as follows:

In our attempt to address these RQs, we analyzed research evidence on these topics by employing a bibliometric analysis of 135 English-language publications from 1990 to 2022 taken from the Scopus database and Google Scholar. The purpose of this analysis was to understand, in-depth, the patterns, methodologies, theoretical foundations, top journals, prominent countries and specific topics in this research area (Paul et al., 2021). We utilized the VOSviewer software version 1.6.18 to create and develop the bibliometric linkages (Van Eck and Waltman, 2014; Paul and Criado, 2020). To advance the SCM literature, this study's systematic literature review (SLR) examined the body of research on artificial knowledge and digitization in SCM through the themes of accountability and sustainable performance for the achievement of the UN 2030 Agenda.

The remainder of this study is structured in six sections. Section 2 comprehensively reviews the theoretical bases of artificial knowledge, digitalization and accountability from the UN's SDG perspective. Section 3 explains the methods and data analysis approaches applied in the study. Section 4 presents the analysis results. Section 5 discusses the findings and describes their implications for theory, practice and future studies. The sixth and last section addresses the study's limitations and offers conclusions.

Artificial knowledge is considered the foremost area in knowledge management, which gains new principles, compiles organizational knowledge, and revolutionizes a firm into a “knowledge organization” in the digital transformation period. Among the core knowledge management areas are knowledge acquisition and interpretation. In terms of artificial knowledge, AI drives the basic principles of fostering the acquisition and interpretation of digital knowledge flows (Stella et al., 2022), including in micro-, small-, and medium-scale businesses (Kumar et al., 2022).

The concept of artificial knowledge in SCM has no universal definition and is still in the emerging and developing stage (Kayikci, 2018). Some scholars have nonetheless tried to define it; for instance, Büyüközkan and Göçer (2018) described the digital supply chain as a value-driven digital system that introduces new methods, latest technologies, and digital analytics into SCM, thereby creating new revenue streams to strengthen business models. Building a digital platform and incorporating data analytics in SCM can maximize value and bring digital knowledge from a variety of sources (Schilling and Seuring, 2021). The advent and outcomes of artificial knowledge in the twenty-first century have fostered various potential developments and comprehensive evaluations in different sectors (Vinuesa et al., 2020). As a result, business organizations now face increasing stakeholder pressure to address digital challenges and improve business operations through digital knowledge and innovations, so as to preserve the integrity of the ecosystem through the digital knowledge management system (KMS) (Joyce and Paquin, 2016; Martins et al., 2019).

Over the last decade, countries have increasingly adopted the UN 2030 Agenda and aligned their business priorities with its global SDGs (de Paula Arruda Filho, 2017). To this end, several knowledge flow methods (Hendriks and Vriens, 1999) have been applied in knowledge acquisition techniques to obtain tacit digital knowledge and expert intelligence systems from domain experts. These techniques are formally functional as they expand the knowledge databases of KMSs and formally document online information (Cherian and Arun, 2022). In addition, multiple knowledge discovery approaches, such as AI-related methods, are effective in identifying interlinkages and trends in knowledge databases to create new digital intelligence (Del Giudice et al., 2020). To promote digital knowledge in such databases, various taxonomies and knowledge maps are often formed as strong foundations for the construction of the databases (Queiroz et al., 2021). In this regard, the implementation of artificial knowledge in knowledge management helps encode digital information in KMSs. For example, multiple AI methods, such as intelligent agents, are applicable to support knowledge search and retrieval techniques in KMSs.

Both the breadth and complexity of resilience theory are relevant in understanding its role, especially in the UN 2030 Agenda (Sullivan and Wamba, 2022). Notably, the resilience theory underpins the theoretical foundation of the artificial knowledge concept. Events like the COVID-19 crisis have catalyzed firms' digital transformation in their processes and structures. In such crisis situations, the challenges in finding new resources and capabilities highlight the need for resilience among institutions, organizations, and individuals (Faruquee et al., 2021). Unexpected crises like the pandemic give managers the chance to analyze their SCM and identify the causes of its disruptions (Fosso Wamba et al., 2022). Resilience, in this context, is the capacity of organizations to take a proactive attitude towards supply chain disruptions and subsequently overcome them to recover balance (Sullivan and Wamba, 2022). Indeed, SCM under extreme conditions is an important topic in the literature (Sodhi and Tang, 2021), which can be intercepted by its relationship with AI (Dohale et al., 2022). Algorithmic fairness is evaluated as much in socio-technical issues (Dolata et al., 2022) as it is in business analytics (De-Arteaga et al., 2022).

A resilient business model represents business organizations' operational capacity to quickly predict, adapt to, respond to, and recover from an unpredictable disruption (Herold et al., 2021). This business model protects against unforeseen events that jeopardize sustainability; thus, it offers several significant strategic perspectives for digitalization and sustainable development. Predominantly, it advocates exchange relationships with different peers, which stimulates peer-to-peer knowledge transfer and acts as a palliative mechanism to build digital knowledge (i.e. artificial knowledge) for stronger digitization as well as to prepare for future crises that threaten sustainable development (Wamba et al., 2017).

Several researchers have pointed out that resilience is among the UN's drivers of SDG achievement, claiming that for a country to be “sustainable,” it needs to work according to “sustainable, resilient, and inclusive principles” (Di Vaio et al., 2021). Considering technologies as enablers of sustainability goals, the role of artificial knowledge in a resilient and sustainable business model indicates that digitalization is able to address stakeholder concerns and respond to external pressure (Modgil et al., 2021; Latif et al., 2022). Indeed, resilience theory allows a better analysis of the linkage between digitalization, artificial knowledge, and accountability for sustainable supply chain performance, which is a neglected aspect in the literature. Among the few studies in this area, Novak et al. (2021) compared equilibrium-based SCM resilience with the view of SCM as a complex and adaptive mechanism.

The resilience lens guides firms' need to strengthen digital platforms to maintain legitimacy in their sustainability behavior. Firms also have to be transparent about their accountability actions to answer stakeholder judgments and meet growing market demands. Stakeholder concerns surrounding digital transformation have contributed to the adoption of digital knowledge in SCM operations. Digital transformation significantly improves the data analytics and data information systems that a business organization provides to its stakeholders, ultimately determining whether the digital platforms can enhance knowledge, resources, and manpower (Wamba et al., 2017). Therefore, according to Sullivan and Wamba (2022), AI is a tool for resilience in firm strategies to rethink SCM as a response to disruptive events; in other words, AI supports the identification of the organizational resources that enable supply chain redesign during disruption management. Consequently, keeping operating processes running improves performance, as these processes are the pillars for the creation of knowledge from AI in the supply chain. This calls for firms to design business models from the perspective of resilience and its supply chain linkages because AI facilitates efficient disruption management and increases performance in the UN's SDGs. Moreover, artificial knowledge should lead to an improvement in SCM accountability. However, despite the enormous attention given to digital technology and knowledge management in the SCM literature (Capestro and Kinkel, 2020; Kamble et al., 2020; Tönnissen and Teuteberg, 2020; Wamba and Queiroz, 2020), scholars have provided scant and inconclusive information on artificial knowledge's implications for future industries, as well as its influences on accountability, traceability, and fraud prevention.

From the resilience perspective, the UN 2030 Agenda does not only encourage firms to invest in artificial knowledge, digitalization, and innovation for long-term planning, but also fosters the resilience business model to enable the greater participation of concerned stakeholders (Tortorella et al., 2023). Stakeholder participation will improve firms' resilience business model and digital platforms, as stakeholders would enforce better management strategies and supervise resilience levels in high-risk events (Nica, 2019). Hence, a resilience business model often corresponds to the number of business partners involved and can enhance firms' sustainable performance, network flexibility, and coping ability against various market fluctuations (Belhadi et al., 2022).

Various approaches have outlined, through different theoretical lenses, how to understand artificial knowledge's role in the development of the resilience business model and how artificial knowledge and digitalization contribute to better sustainable performance (see Figure 1). For example, institutional theory explains the relationship of artificial knowledge with the resilience business model by stating that the use of modern digitalization addresses stakeholder concerns and responds to external pressure. Specifically, artificial knowledge is one of the various innovative channels organizations use to modify the institutional frameworks that transform and promote digital platforms (Bag et al., 2021; Hinings et al., 2018). Institutional theory has a sociology aspect, wherein legitimacy basically defines managerial decisions. On the other hand, its economic aspect asserts firms' desire to accomplish isomorphism by increasing productivity.

The literature has also discussed the prominent role of accountability in attaining a competitive advantage when it is fully integrated into a firm's operations in a unique and irreplaceable manner (Barney, 1991; Kozanoglu and Abedin, 2020). According to Logsdon and Lewellyn (2000), stakeholder accountability can be a key success factor for the corporate accountability process. Correspondingly, the legitimacy theory argues that objectionable and inappropriate accountability behavior is exposed by larger legitimacy forces and creates incentives to improve sustainable performance (Cormier and Magnan, 2015). The literature on sustainable business legitimacy recommends that the achievement of legitimacy be based entirely on the beneficial results of accountability (Tilling, 2004) and explicit moral discourse on a firm's acceptability. Finally, based on legitimacy theory, Mobus (2005) stated that those organizations that conduct their accountability practices in accordance with social values and norms achieve greater legitimacy.

As opposed to a narrative review, the approach we used to analyze the relevant literature in this study was the SLR, which utilizes a scientific, transparent, repeatable procedure (Treanfield et al., 2003; Snyder, 2019). It is also a well-established approach in SCM research (Durach et al., 2017; Snyder, 2019; Donthu et al., 2021; Lim et al., 2022). The SLR enables the systematization and classification of key findings in the research area, emphasizing unknown characteristics so as to develop directions for future research (Kraus et al., 2022; Martins et al., 2019; Paul et al., 2021; Di Vaio et al., 2022b). The advantages of SLR analysis include: better result quality (Christofi et al., 2017); minimization of distortions (Dada, 2018); greater validity and replicability (Wang and Chugh, 2014); a clearer roadmap for the field under study (Kauppi et al., 2018); and the prediction of various factors that build a novel conceptual framework as future research agenda (Dada, 2018; Lim et al., 2022). Specifically, our SLR analysis progresses the research fields of artificial knowledge, digitization, and the accountability-based business model (Donthu et al., 2021).

The methodology of this study consisted of two distinct research phases: a) identifying, reading, and interpreting pertinent publications; and b) performing a bibliometric evaluation of the selected papers. In line with the procedure recommended by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method, we conducted four steps in the initial phase. They were: (1) the identification of published papers from repositories; (2) the screening of the papers; (3) the selection of relevant papers based on eligibility; and (4) the finalization and inclusion of the papers for analysis. Figure 2 illustrates the data collection and analysis procedure followed at each level of this study to ensure a trustworthy methodology.

We began with the database selection in the first phase. In accordance with Fink (2019), the selection procedure culminated with choosing the Scopus and Google Scholar databases. Scopus is the most comprehensive abstract and citation database available to scholars, government institutions, and business organizations (Fahim and Mahadi, 2022). It has over 1.8 billion cited sources from as far back as the 1970s. It encompasses 84 million records from seven thousand publishers, 17.6 million author profiles, and nearly ninety-five thousand affiliation profiles. It is also useful on account of its h-index, a score that reflects the quality of an article, author, or journal. We opted to combine the results of our Scopus database search with that of our manual search on the Google Scholar site to improve the scope of the selected topic, since Scopus has greater coverage compared to the Web of Science (WoS) database. Previous researchers have also used a similar strategy (see Di Vaio et al., 2022c).

To collect all the works on artificial knowledge, SCM digitalization, and accountability, we picked a broad period of study spanning over 30 years from 1990 to 2022. However, the database search revealed that the initial article on this topic was only published in 2004 (see the following section for more details on the year-wise publication record). Furthermore, to discover all relevant papers, we executed numerous search queries via shortened (truncated) associations between nine search string categories, as mentioned below:

1. Group 1: artificial knowledge AND digitalization AND SCM

2. Group 2: digital transformation AND resilience AND SC

3. Group 3: artificial knowledge AND digitalization AND SCM AND accountability

4. Group 4: artificial knowledge AND digitalization AND SCM AND accountability AND sustainable performance

5. Group 5: digital transformation AND resilience AND SCM AND accountability

6. Group 6: digital transformation AND resilience AND SCM AND sustainable performance

Following earlier works (e.g. Di Vaio et al., 2022b), the selection of relevant publications focused on the above-combined categories to highlight probable connections between the results obtained in the six groups, rather than discarding notable contributions to the problem being examined. The specific keywords used in the first stage in combination with the research theme included “artificial knowledge”, “digitalization”, “supply chain management (SCM)”, “digital transformation”, “resilience”, “accountability” and “sustainable performance”. Digitalization was often inserted in the search since it was the focal point of our study. By searching for these terms in the articles' title, abstract, or keywords, the linkages between SCM digitalization, accountability, and sustainable performance were identified. The initial keyword exploration yielded 607 articles, which were then filtered for journal papers in the English language on the selected research topics (e.g. social sciences, computer sciences, business management, and decision sciences). This reduced the number of articles to 178.

In the second stage, relevant publications were chosen based on our research criteria and content analysis of the articles' abstracts. Closely analyzing the abstracts' content enabled us to arrange the data in a repeatable fashion, and subsequently, emphasize the relevance of each article to the themes presented in our study. Taking into account the possibility that the Scopus database does not contain all existing articles pertaining to this study, we also manually searched for articles on Google Scholar, employing identical search parameters to explore prominent journals known to publish articles on artificial knowledge in SCM digitalization, accountability, and sustainable performance measures. Specifically, we focused on journals such as Annals of Operations Research, International Journal of Productivity and Performance Management, Journal of Enterprise Information Management, Supply Chain Management, Journal of Business Research, and International Journal of Supply Chain Management to avoid any exclusion of papers relevant to our study's objectives.

The final phase of this study centered around each scholarly work, wherein we thoroughly examined every paper to determine important areas related to the themes under study. We reviewed our data and removed duplications and extraneous articles, yielding a final list of 135 articles (see  Appendix for summary of selected articles). Then, we began performing the various parts of the SLR by feeding the final article list into the VOSviewer software (version 1.6.18), a free computer program used to create, visualize, and explore network data maps (Van Eck and Waltman, 2017). The development of a mapped representation of bibliographic data in the VOSviewer facilitates a broader and more exact comprehension of the impacts of the research topics (Di Vaio et al., 2022c). Moreover, in its 2014 update, the VOSviewer incorporated extensive text-mining capabilities, using which two-dimensional term maps can be generated from a collection of texts to reflect terms' relationships based on location and distance. Notably, the correlations between terms are defined by their co-occurrence in the articles (Van Eck and Waltman, 2017). Additionally, to support citation analysis, Harzing's “Publish or Perish” (POP) software was utilized in this study. It is, again, a free software that allows users to retrieve and analyze academic citations (Jacsó, 2009). The findings of the analyses described above are detailed in the following section.

Bibliometrics was utilized to conduct advanced statistical and graphical categorized tests that summarized the articles' data and highlighted its spatio-temporal aspects. Indeed, bibliometric analysis results in more trustworthy and systematic results about a chosen topic without the possibility of neglecting prior works (Di Vaio et al., 2022c).

Using the bibliometric analysis, we produced a conceptual map depicting the interrelatedness among the keywords included in the database search. The overlay depiction of the keywords, as categorized by color matching, is shown in Figure 3. The figure displays the color relationships, which are calculated as the frequency index of word recurrence throughout time. It is noteworthy that the terms “sustainability”, “supply chain”, and “industry 4.0 model” are correlated in terms of their color correspondence. Similar linkages can be seen between “SCM and digitalization”. Unsurprisingly, “COVID-19” emerged as an independent, commonly used keyword in the relatively recent literature.

Table 1 lists the most popular keywords used by previous authors. The statistics showed that “Blockchain” (n = 22, 4.31%), “sustainability” (n = 21, 4.11%), “supply chain” (n = 19, 3.72%), “SCM” (n = 16, 3.13%), “COVID-19” (n = 12, 2.35%), and “industry 4.0” (n = 12, 2.35%) are the top six terms.

The publishing history of the articles on the chosen topics from January 2004 to June 2022 is depicted in Figure 4. Only four papers were published between 2004 and 2016, with one each in 2004, 2006, 2010, and 2014. As a result, the identified publications in this subject area over the past decade represent somewhat less than 3% of the overall articles published in the area (see Table 2). Since 2017, the number of articles published on AI, digitalization, supply chains, and accountability has steadily increased. This demonstrates academics' growing global interest in AI, digitalization, supply chain, and accountability research. The number of papers published in this field reached a peak in 2021. The two-year moving average plotted in the dashed line in Figure 3, on the other hand, indicates that more articles will be published in 2022 than 2021, contributing to the continuation of the increasing trend seen since 2017. Table 2 contains the whole year-by-year list of articles.

Table 3 lists journals that have published at least two articles. In this category, the famous journals were as follows: International Journal of Supply Chain Management”, “Annals of Operations Research”, “International Journal of Productivity and Performance Management”, and “Journal of Enterprise Information Management”. The analysis of these journals, as seen in Table 3, suggests that most of the published articles were concentrated in operations management and SCM journals.

Conversely, as per Table 3, more than 51% of the papers have been published in various other journals from different disciplines. The  Appendix shows the complete list of these journals. This is indicative of the diversity of journals in which SCM digitalization and accountability papers have been published.

Table 4 categorizes the publications based on their broad subject areas. Evidently, the “Business, Management and Accounting” subject area has witnessed the highest number of publications, followed by the “Computer Science” and “Decision Sciences” subject areas. The variety of journals and their subject areas indicate the diversity of the functional disciplines these publications come from.

More specifically, the selected journal papers have addressed a range of topics, including the sustainability of logistics (Rahman et al., 2019), development of virtual relations (Ukolov et al., 2019), digital SCM and development (Afanasyev et al., 2019), digital economy development (Kartskhiya et al., 2020), digitalization of energy manufacture (Afanasyev et al., 2019), sustainable supply chain finance (Reza-Gharehbagh et al., 2022b), AI and blockchain adoption (Chatterjee et al., 2021; Vafadarnikjoo et al., 2021; Grover, 2022), AI-driven innovation (Belhadi et al., 2021), sustainable trade promotion (Wu et al., 2020), epidemic outbreaks in supply chains (Queiroz et al., 2020), and the circular economy of industry 4.0 (Lopes de Sousa Jabbour et al., 2018).

The percentages of publication contributions by country are displayed in Table 5. With 23 publications (16.55%), the United States tops the list, followed by India with 22 publications (15.83%). This finding shows that the early epicenters of SCM digitalization research were the United States and India. Unexpectedly, Australia comes in eighth place (n = 7, 5.04%), deviating from the findings of Di Vaio et al.’s (2022c) SLR on the contribution of blockchain technology to gender equality, in which Australia was ranked second. The United States has a strong growth rate and promotes digitization, supply chain, AI, and accountability thanks to the government's effective efforts in these areas (Thylin and Duarte, 2019; Di Vaio et al., 2022a). To ascertain the drivers of the growing digitization of the United States, an extensive study on AI, digitalization, supply chains, and accountability is being carried out.

Table 5, on the other hand, reveals that the publications come from just 10 different nations. This statistic draws attention to the fact that most research in this area is concentrated in a few countries, underscoring the necessity of studies with a larger geographic scope to depict the global picture of supply chain digitalization and accountability. On the world map, Figure 5 illustrates the geographical spread of these nations. Interestingly, there appears to be no research from the continents of South America or Africa. The countries in these continents are a part of the lower tiers of many global supply chains. Unfortunately, most of these nations are the ones that fall behind in terms of accountability and digitalization; the paucity of publications resonates the same shortcoming. The map also shows that Asia dominates this field of study.

The foremost prolific authors in the domains of AI, digitalization, supply chains, and accountability are listed in Table 6. With three publications apiece, Joshi S (India), Kumar A (United Kingdom), and Sharma M.R. (United Kingdom) lead the list. Belhadi A., Gunasekaran A., Kumar M., Tsolakis N., and Mani V. are also active authors, each with two articles and more than 30 citations. It is noteworthy that these authors represent a variety of genders and are from both developed and developing nations.

Academic cooperation is essential for the advancement of any subject; hence, increased international collaboration is necessary (Turner and Baker, 2020). Figures 6 and 7 demonstrate the level of interaction among academics using nations and individual scholars as units of analysis. The United States, India, the United Kingdom, and the Russian Federation are the nations with the highest authority in collaborative projects. It is also interesting to note the significant collaborative publications between India and the United Kingdom due to the network between several key researchers, namely Joshi S (India), Kumar A (United Kingdom), and Sharma M.R. (United Kingdom). This study thus reveals that a strong network of collaboration exists across all continents. Individually, the most prominent writers are Pumputiene, E., Pozzi, M., Pauschinger, T., Morgione, S., and Rossi, S., who have collaborated on multiple publications.

In research in general, increasing collaboration among scholars from various countries is observed. Cultural affinities, language, and geographical location, are all determinants and drivers of co-authorship decisions (Di Vaio et al., 2022a). It is revealed that not only do the United States and the United Kingdom publish more research articles, but their academics also work more successfully with their peers in other countries. Perhaps, this success is related to the national emphasis on digitization and accountability in these countries.

Table 7 lists the top institutions that have published at least three articles on AI, digitalization, supply chains, and accountability. Four articles have been contributed by the Università degli Studi di Napoli Federico II in Italy and the Montpellier Business School in France.

Figure 8 depicts a network visualization map of the affiliation of co-authors. The most active institutions in this field of study are the Montpellier Business School in France and Cadi Ayyad University in Morocco.

Citations are used in research assessment to show how much a publication has drawn on other publications' ideas, research, and content. As a result, the impact of a study is determined by the number of citations it attains (Bornmann and Daniel, 2007). Table 8 shows the most frequently cited scholars and articles. The study entitled “Firm performance impacts of digitally enabled supply chain integration capabilities” authored by Rai et al. (2006) received the most citations in the selected database. Lopes de Sousa Jabbour et al.’s (2018) published work titled “Industry 4.0 and the circular economy: a proposed research agenda and original roadmap for sustainable operations” is the second most prominent article. The last column of Table 8 indicates that the quantity of citations generated every year by the “Publish or Perish” program is based on citation numbers from Google Scholar. Accordingly, based on Table 8, the most prominent paper by citation number is Queiroz et al.’s (2020) “Impacts of epidemic outbreaks on supply chains: mapping a research agenda amid the COVID-19 pandemic through a structured literature review.” The prominence of the study is mainly because of its highlight attribution (COVID-19 pandemic) and nature (literature review).

VOSviewer provides a valuable tool to cluster selected publications and analyze the resulting clusters in aggregated graphical form (Van Eck and Waltman, 2017). Accordingly, it enables researchers to identify significant keywords in articles within a cluster, as well as the co-occurrence frequency among them. According to Bornmann et al. (2018), nodes in the co-occurrence map show the correlations between two terms. Following Di Vaio et al. (2022b), in this study, we selected 16 keywords (specifying the minimum number of keyword occurrences to six) and assessed the intensity of co-occurrence links with other keywords. Subsequently, we identified two separate clusters, distinguished by color in green and purple, respectively.

Figures 9 and 10 depict the co-occurrence of keywords in all fields in a graphical format. The VOSviewer generates different circles and colors upon analyzing terms. Circle size is an indication of how frequently a keyword appears in the selected field, while the space between circles implies the linkage between keywords (i.e. the greater the space, the weaker the connection between the keywords) (Van Eck and Waltman, 2014; Di Vaio et al., 2022b). In our analysis, the software generated five clusters with different colors (i.e. red, blue, green, purple, and yellow) depending on their associations and fields when fractional counting was used (see Figure 9). However, when full counting was used, it generated only four clusters in red, blue, green, and yellow (see Figure 10). Our identification of the two distinct clusters in these figures was based on the highest number identified by the software for each cluster. These figures suggest that in SCM, digitalization and artificial knowledge occur through the adoption of blockchain, AI, and digital storage. These could lead to improved decision-making towards sustainable development. Therefore, governments, policymakers, and business organizations should aim to strengthen the use of advanced digital platforms and technologies to achieve improved accountability and sustainable performance. Moreover, incorporating these technologies can build resilience in the face of unanticipated events that endanger sustainable growth, such as COVID-19, by boosting a company's capacity to predict, adapt, respond, and recover rapidly (Herold et al., 2021).

We performed content analysis of the articles in tabular form, as presented in  Appendix. The analysis provided a brief profile of each article in terms of its aims, findings, methodologies, and underpinning theories. Scholars have examined the use of digitalization and artificial knowledge in SCM in various industries, such as food and agriculture, ICT, fashion, logistics, marine, textile, tourism, healthcare, fishing, automotive, pharmaceutical, construction, minerals, mining, and manufacturing (Bechtsis et al., 2017; Cherviakova and Cherviakova, 2018; Garcia-Muiña et al., 2018; Pongpanit and Sornsaruht, 2019; Alharthi et al., 2020; Calvão and Archer, 2021; Chen et al., 2022; Griffin et al., 2022; Joshi and Sharma, 2022; Mahroof et al., 2022; Mishra et al., 2022; Oguntegbe et al., 2022; Shamout et al., 2022; Sharma et al., 2022; Shi et al., 2022), and across multiple levels including SMEs, cities, individuals, the government, and the world (Babenko et al., 2020; Bellingan et al., 2020; Wong et al., 2020; Bagloee et al., 2021; Bisogni et al., 2021; Hjaltadóttir and Hild, 2021; Potocka-Sionek, 2021; Nasir et al., 2022; Nudurupati et al., 2022; Reza-Gharehbagh et al., 2022a).

The findings from our analysis highlight the wide interest of scholars in investigating digitalization and AI issues in SCM. Particularly, regarding RQ1 “What is artificial knowledge in the digitization of SCM?”, our results assert the increasing significance of artificial knowledge and the resulting need to extend the focus of artificial knowledge and digitization research to supply chain operations. Under the theoretical lens of resilience, SCM has to be rethought and transformed from a reactive approach into a proactive approach. Consistent with this, Sullivan and Wamba (2022) found in their research that AI can be firms' reply to disruptive events like the COVID-19 pandemic. Moreover, based on our analysis results, the progress of the resilience business model highlights the substantial necessity to consider cultural and social outcomes in addition to the financial outcomes of artificial knowledge adoption. From this viewpoint, this study advances existing knowledge on the need to develop a resilience business model for artificial knowledge and digitalization in SCM (Queiroz and Wamba, 2019).

Additionally, according to the results of this study, the engagement of communities should be included in the strategic rethinking of SCM (Song et al., 2022). This means that the resilience business model in SCM does not have to be limited to the adoption of AI merely as a response tool to crises to reduce negative impacts on operational performance; rather, AI in SCM resilience should also drive the social pillars of sustainability. The results thus explain that artificial knowledge currently attracts substantial attention not only to achieve economic goals but also to promote the well-being of cultural and social communities.

Apart from being a growing concern, the usage of intelligent systems and advanced digitalization is recognized as a revolutionary way for modern businesses to strengthen artificial knowledge. Indeed, the emergence of digitalization is considered a major contributor to the implementation of artificial knowledge. In this regard, organizations should focus on more advanced digital platforms and engage in an advanced holistic approach that fosters the organizational structure by increasing the reach, precision, and speed of digital platforms and information processing systems (Soto-Acosta et al., 2018; Wirtz et al., 2019). The establishment of digitalization entities can further result in substantial speed and quality improvements in data analytics and information processing systems, which offers greater access to individuals. Subsequently, by using artificial knowledge effectively and efficiently, organizations can facilitate digital learning and use resources in a better manner.

Regarding the first part of RQ2 “Does the digitalization of SCM increase accountability?”, the results highlight the importance of the resilience business model in the SCM literature by pinpointing accountability as a critical driver of artificial knowledge and digitalization. This encourages organizations to improve their innovation more transparently, thus improving sustainability. The accountability mechanism encourages new investments in artificial knowledge and digital technology-related initiatives to reinforce digital technology implementation in supply chain production systems. It also strengthens the information system at all levels, promotes artificial knowledge in the workforce, and leverages and improves data analysis (Warner and Wäger, 2019).

The second part of RQ2 “Does the digitalization of SCM increase sustainable performance?” was answered by our analysis, which highlighted the significance of artificial knowledge being promoted by multiple societal actors including institutions, organizations, and civil communities. In addition, artificial knowledge drives the organization towards a sustainable development strategy and provides the digital transformation required to strengthen sustainable business models. It further encourages firms to invest more in technology-oriented development by partnering with other companies, subsequently advancing their sustainable development agenda. Artificial knowledge is also recognized as a key element for businesses that are not only spreading digital transformation, but also transforming digital knowledge into business processes and incorporating digital technologies and novel sustainable solutions to achieve the SDGs. In this regard, artificial knowledge is specifically aimed at bringing digital transformation and ensuring various sustainable business models for supply chain operations. It does so by serving as a mechanism of production and consumption as well as by promoting digital knowledge in SCM. Fosso Wamba's (2022) study reported on the significance of AI integration in all phases of firms' operational processes to “create and capture” AI value. The results of our study provide another dimension to the value derived from AI, namely artificial knowledge.

While the prior literature affirms that digital transformation is a supporting mechanism for capabilities (Matarazzo et al., 2021; Verhoef et al., 2021), its association with the resilience perspective has been paid scant attention. In line with previous research (Mikalef and Gupta, 2021), our present findings evince the need to prioritize the role of IT capabilities in improving capabilities like culture, management, employees, and infrastructure (Gong and Ribiere, 2021), based on the resilience approach in managing artificial knowledge in SCM. In fact, the role of resilience is key in contributing to the sustainable capabilities of firms. In previous literature, IT capabilities were conceptualized into three dimensions, i.e. infrastructure, business expansion, and proactive stance. However, each of the IT capabilities listed above have rarely been considered from the resilience perspective. If a firm wants to achieve a competitive advantage, IT capabilities must adopt the effective resilience approach to reach a breakthrough. In this regard, opening up to the new avenue of resilience can develop an organization's socio-environmental well-being by advancing artificial knowledge and strengthening IT capabilities (Liebowitz, 2001; Buzko et al., 2016; Jia et al., 2018; Haseeb et al., 2019; Wiesboeck et al., 2020).

Ultimately, this analysis raises interesting perspectives; nonetheless, there remains the matter of identifying a point of interconnection. Digital and sustainable challenges can travel on different tracks; similarly, the varying interests of stakeholders may not coincide with those of the next generation (D'Adamo and Gastaldi, 2022). Indeed, businesses have as their goal the well-being of customers, but also seek to maintain a balance of eco-systems by focusing on resilience as a strength and not as an inability to react to market shocks and changes. Digital knowledge, when combined with human knowledge, can make the internal environment more conciliatory and more responsive to changes in the external environment. The circularity of resources allows for the optimization of the production process (Taddei et al., 2022), and pushes accountability towards green choices that may or may not be recognized by consumers (Liu et al., 2022).

Artificial knowledge is applied in several different sectors with different user intensities, such as AI-based systems, knowledge-based IT systems, and intelligent executive systems. In today's modern digital world, artificial knowledge has risen as a favorable and popular form of digital transformation, helping organizations promote advanced algorithms and Big Data analytics. As we concluded, the use of artificial knowledge is attractive as it can affect decision-making abilities and expand IT strategies. Our study elaborates on the role of artificial knowledge and digitalization in SCM to increase accountability and sustainable performance, unfolding aspects of artificial knowledge and digitalization in various ways. First, this study addresses the role of artificial knowledge in the resilience business model in SCM to achieve the UN 2030 Agenda and its SDGs. Second, the current study points out that the accountability mechanism is critical for artificial knowledge and digitalization to facilitate better decision-making, create value, and achieve sustainable performance. Third, it confirms the need for application cases that did not emerge from this analysis (Negri et al., 2021).

The present study's findings offer valuable information and guidance to supply chain managers. First, the importance of artificial knowledge and digitalization has various implications for resource selection, employee skill development, and Big Data-oriented culture creation. Specifically, IT leaders and professionals should institute appropriate artificial knowledge and digitalization practices that can shape resource selection strategies within their organizations (Soto-Acosta et al., 2018). Moreover, training for digital transformation and education programs should include cultural change towards accountability in SCM, so as to increase sustainable performance in sustainable and resilient business models. In addition, our findings serve as a guideline for Big Data practitioners by emphasizing that building artificial knowledge and digitalization to fulfill various demands requires not only financial investments but also adequate intangible assets such as time, effort, and human skills. Perhaps the most significant result of this research is the useful insight into artificial knowledge's dynamic role in decision-making. Consequently, all efforts should be put towards responsible and ethical AI governance and its benefits to improve firm performance (Papagiannidis et al., 2022). Our study has looked into scholars' and practitioners' perceptions of artificial knowledge and digitalization to make better decisions in SCM operations. Hence, this study provides a new awareness of artificial knowledge and digitalization, extending prior research that has mainly presented organizations' application of strategies and techniques for new technologies. Moreover, this study points out how the digital transition cannot transcend the sustainable transition in enforcing suitable solutions to external changes.

Artificial knowledge can play a critical role in policymaking. In developing an effective digital KMS to achieve resilient and sustainable business models in SCM, policymakers should focus on critical technology drivers that forecast cultural drift. In doing so, policymakers and practitioners should establish the missing link of “effective technology-oriented policies” to promote digital KMSs whose goals are aligned with AI technology and the UN 2030 Agenda. It is relevant for policymakers to design the latest models of AI technology which can provide daily updates in addressing digital challenges with minimal resources. Finally, the use of public funds should reward territorial realities of excellence that identify the models to be emulated, such that replicability is made possible.

The literature has mainly scrutinized the relevance of artificial knowledge and digitalization with the aim of exploring how organizations can improve the accountability process and succeed in achieving sustainable SCM performance. Based on our SLR in this paper, we offer the following propositions (see Figure 11):

Scholars have addressed the need to explore and implement artificial knowledge and digitalization as an agenda for improving SCM operations (Sambasivan et al., 2009; Samuel et al., 2011; Gansser and Reich, 2021; Del Giudice et al., 2023; Chen et al., 2022; Liu et al., 2013). Nonetheless, the role of artificial knowledge in SCM operations has not been sufficiently addressed (Gold and Heikkurinen, 2018). Furthermore, various studies have neglected to examine the impact of artificial knowledge with regard to the digitalization of SCM operations (Chen et al., 2022). Therefore, it is beneficial for future research to pay attention to artificial knowledge in the digitalization of SCM operations. In addition, multiple knowledge discovery approaches, such as AI-related methods, can detect linkages and trends in knowledge databases for the creation of new digital information in SCM operations. To promote digital knowledge, various taxonomies and new learning maps are often formed as strong foundations for the development of such databases. The implementation of artificial knowledge in the field of SCM helps encode digital information in KMSs. Multiple AI approaches, such as intelligent agents, are in fact applicable to support knowledge search and retrieval techniques in KMSs.

This proposition on SCM digitalization and accountability was proffered based on studies that explain digitalization in SCM and in response to our RQ2. Multiple studies have highlighted the significance of digitalization in SCM, particularly in innovating various accountability mechanisms and changing conventional accountability frameworks into more transparent and sustainable ones. By adapting advanced technologies, SCM operations gain improved traceability and value chains, which achieve better accountability. According to Logsdon and Lewellyn (2000), digitalization can be a key success factor for the accountability process. Furthermore, digitalization has exposed questionable and inappropriate accountability behaviors through broader legitimacy forces and created incentives to improve accountability mechanisms. The literature recommends that the achievement of the accountability mechanism is based entirely on the beneficial results of digitization (Tilling, 2004) as well as the explicit moral discourse on the acceptability of any company. Also, with regards to SCM operations, organizations that promote digitalization platforms in accordance with social values and norms should achieve greater accountability. This can be accomplished by enriching and augmenting organizational knowledge (Harfouche et al., 2022) and assigning greater importance to behavioral considerations (Pournader et al., 2021).

Our final proposition also answers RQ2 and is driven by research on the way digitalization facilitates the assimilation of resilient and sustainable business models in SCM towards achieving the SDGs. Currently, SCM is under increasing pressure from stakeholders to counter digital challenges and improve business operations by addressing digitalization so that superior sustainable performance can be achieved (Joyce and Paquin, 2016). By enabling digitalization to advance in SCM, for example through AI, new digital knowledge can be created based on the identification of trends and linkages in knowledge databases. In turn, the various taxonomies and new knowledge maps formed to promote digital knowledge in such databases act as the foundations for the construction of a sustainable organization. Implementing artificial knowledge in SCM encodes digitization and supports SCM's knowledge promotion and recovery methods for sustainable performance. Notably, this can be achieved if shared value is associated with all stakeholders (Appolloni et al., 2022).

Artificial knowledge and digitization drive incredible transformations in SCM by driving new ways and best practices for organizations to interact with various digital platforms. When promoted and shared, artificial knowledge and digitization can yield substantial benefits alongside digital information interpreted within the network. This study highlights that if well applied, artificial knowledge can also achieve the UN's SDGs by strengthening decision-making and accountability. From here, the first limitation of this work emerges; given that it is an SLR, there is a need to verify the propositions put forth in this study with actual application cases.

Taking into account the converging characteristics of artificial knowledge and digitization in various bodies of academic literature, this study systematically reviewed 135 articles that have analyzed these concepts in the SCM context, as well as the accountability process that can strengthen digital platforms to meet the UN 2030 Agenda. Artificial knowledge and digitization research within SCM provide a detailed description of the present condition of the SCM field and highlight the main critical challenges facing the world today. Artificial knowledge and digitization aim to improve various accountability measures and suggest new business practices to achieve sustainable performance. Studies on these two concepts discuss the current state of digital technological developments and highlight the focus on various technologies. They open a debate on government initiatives for digital business platforms and the implications of the digital revolution. This gives rise to the second limitation of this study, which was constrained to conceptually investigating the use of artificial knowledge in the resilience business model. It might be stimulating to carry out a quantitative exploration of artificial knowledge, specifically on how organizations understand this concept to promote digitization. In conclusion, this study's conceptual framework on artificial knowledge and digitization in SCM aims to increase SCM accountability and sustainable performance, especially when disruptive phenomena or crises occur which require the resilience of business organizations.

The authors would like to thank the Editor-in-Chief, Guest Editors and anonymous referees for providing helpful comments and suggestions, which led to improving the article. This work is an outcome of the “BlueShipping&Cruise Lab” (BSCLab), Department of Law, University of Naples Parthenope, Italy.

Funding: This work was supported by the University of Naples Parthenope, Naples, Italy, Research Financial Resources, “Ministry of University and Research Ministerial Decree of 25.06.2021 n. 737 for research project entitled Digital transition for Sustainable and Resilient Business Models in the ship-port interface towards the 2030 Agenda” – Principal Investigator Prof. Dr. Assunta Di Vaio.

Table A1