Business Model Innovation through AI Adaptation: The Role of Strategic Human Resources Management
[Correction added on 27 January 2025, after first online publication: The affiliations for the author Michael Christofi have been updated and the affiliation for the author Shlomo Tarba is now updated with superscript 8 in this version.]
Abstract
While artificial intelligence (AI) requires business model innovation, it simultaneously poses persistent operational, regulatory and strategic challenges, highlighting the importance of researching AI adaptation to appropriate organizational value. AI adaptation is not monolithic, and its nature and consequent value appropriation processes may vary due to external factors and an organization's strategic approach to innovation and resource management. Accordingly, a taxonomy of AI adaptation and its link with value appropriation can yield a theoretical understanding and practical implications of why and how organizations vary in leveraging strategic human resource management to shape business innovation led by AI adaptation. In this paper, we address this issue by applying adaptive structuration theory and conducting interviews with top management personnel from 51 companies based in India. Based on our findings, we develop a novel taxonomy of AI adaptation (exploitive, exploratory, emancipatory and expedient), structured within a 2 × 2 matrix and a robust model of value appropriation within a dynamic business environment.
Introduction
Artificial intelligence (AI), heralded as a source of value, is increasingly becoming an integral part of business management (Huang and Rust, 2021). Its use transcends sectorial and national borders and has ushered in the rapid development of Industry 4.0 (Rampersad, 2020) and, more recently, Industry 5.0 (Noble et al., 2022). AI is not only a change agent within the organizational structure (e.g. Einola and Khoreva, 2023); it can also pivot transformative and innovative management (Von Krogh, Roberson and Gruber, 2023). Given the growing integration of AI into organizational practices and processes, it is imperative to develop a better comprehension of organizational strategies for using and adapting AI (Bankins et al., 2024).
Due to AI's contribution to innovative business modelling and generating value for various stakeholders – including employees and human resource (HR) managers – its effective design and use have become strategic priorities for organizations (Chowdhury, Budhwar and Wood, 2024; Davenport, 2018). However, the proliferation of the use of AI is accompanied by a persistent debate concerning the structural and processual bottlenecks that impede its appropriate use (Huang and Rust, 2021). A survey of UK-based financial services suggests that the deployment of AI can be hindered by many obstacles, such as difficulties in keeping pace with data growth, challenges with customization and a lack of skilled staff.1 Due to AI's pervasive influence on businesses and their employees, governments and regulatory bodies have developed policy frameworks aimed at forcing organizations to adapt AI creatively and context-specifically (Ameen et al., 2024).
Technology adaptation has gained significant research attention, highlighting the essence of the socially and contextually appropriate use of technology (Dey, Newman and Prendergast, 2011; Kung et al., 2015). Drawing from the information systems (IS) literature (Bala and Venkatesh, 2016; Muhammad et al., 2021), AI adaptation can be defined as employees’ cognitive and behavioural efforts to manage any perceived consequences associated with AI. Although the adaptation of technology in general – and AI in particular – is centred on individual employee behaviour, it is facilitated by organizational strategies and practices (Chowdhury et al., 2022, 2024). In particular, an organization's strategic human resource management (SHRM) influences AI adaptation through the recruitment of new staff members and the retention, training and development of existing ones (Ameen et al., 2024; Chowdhury, Budhwar and Wood, 2024).
The literature identifies two orthogonal types of adaptation in terms of how technology is changed or adjusted to be applied within a given context: exploration and exploitation (Schmitz, Teng and Webb, 2016). Users may wish to explore various applications suited to maximize benefits through radical adaptation. Alternatively, they may adhere to certain application(s) and make incremental changes through exploitation (Shao, Li and Wang, 2022). However, technology adaptation may not be of a binary nature, as users may choose not to use a new technology and/or to revert to an old one (Bala and Venkatesh, 2016). Hence, there is a lack of consensus regarding the nature and types of technology adaptation. This can hinder the conceptualization of how employees utilize different forms of adaptation to create and appropriate value for their organization.
The technology adaptation within an organization is closely connected to the organization's SHRM capabilities and long-term growth strategies (Cozzolino, Verona and Rothaermel, 2018). SHRM is vital because it enhances performance by driving employee engagement with new technology. Additionally, technology adaptation determines how an organization can utilize its internal resources to maximize benefits (Dey et al., 2019), emphasizing the importance of further research on the role of SHRM in creating organizational value through technology integration.
When fostering organizational growth and success, it is important to understand that creating value is not an end in itself (Mizik and Jacobson, 2003). It has been proposed that a part of the value created should also be appropriated within an organization to benefit its internal and external stakeholders for sustainable growth and strategic advancement (Corsaro, 2020). Value appropriation refers to the degree to which a company can capture value during an exchange process (Volschenk, Ungerer and Smit, 2016). In essence, value appropriation, distinct from value creation, represents the net benefits an organization retains when engaging in stakeholder exchanges (Albert-Cromarias, Asselineau and Blanchard, 2022).
- To analyse how and to what extent SHRM can shape AI adaptation and appropriate value within a dynamic business context.
Our conceptual underpinning is built on adaptive structuration theory (DeSanctis and Poole, 1994; Orlikowski, 1991). Adaptive structuration theory (AST) enables us to analyse how the relevant actors and agents within an organization iteratively shape and reshape the use of technology (Shao, Li and Wang, 2022). By undertaking an extensive qualitative enquiry into 51 large and medium-sized organizations based in India, we contribute to the body of knowledge by developing a novel taxonomy of AI adaptation presented through a 2 × 2 matrix combining management innovation and the extent of value appropriation. The matrix enables us to analyse how SHRM can engage in different types of AI adaptation to appropriate value.
In the following section, we present a review of the relevant literature, which enabled us to identify the research gap. Subsequently, we explain our research methodology, present our findings, discuss the findings and draw a conclusion.
Literature review
AST and AI adaptation
Based on the literature on the adaptation of technology within organizational contexts (Muhammad et al., 2024; Schmitz, Teng and Webb, 2016; Shao, Li and Wang, 2022), we adopted AST (DeSanctis and Poole, 1994) as the conceptual lens for our study. AST enables us to analyse the nature and types of technology adaptation and understand how an organization's internal processes and resources (e.g. human resource management, HRM) shape and reshape technology adaptation to achieve the desired outcome (appropriated value).
DeSanctis and Poole (1994) introduced the concept of AST to explain technology adaptation within group decision support systems. AST is focused on the role played by technology in creating and/or recreating social structures. Human agents use their knowledge, ability and interpretation to reciprocally and iteratively engage with technology (Orlikowski, 1992). Human agency is considered a knowledgeable and voluntarist entity that often embodies the desire to create new structures (Kallinikos, Aaltonen and Marton, 2013). Accordingly, within the context of the use of technology in an organization, some employees accept the existing structures. In contrast, others seek to amend or replace them by applying their knowledge and ability (Bernardi, 2017). Employee interaction with technology within an organizational context has significant implications for personal development, performance and satisfaction (Huang and Rust, 2021; Vrontis et al., 2022). Accordingly, AST is a suitable theory to enable us to unpack the inter-relationship between technology adaptation and SHRM within an organization.
Based on AST, Schmitz, Teng and Webb (2016) conceptualized the ‘exploitive’ and ‘exploratory’ uses of technology, which endure as meta-interpretations of adaptation, forming an orthogonal duality (Schmitz, Teng and Webb, 2016) and reflecting a nuanced interpretation of Tyre and Orlikowski's (1994) assertion that adaptation behaviours can be incremental/gradual or transformative/abrupt. However, Schmitz, Teng and Webb's (2016) classification falls short of considering employee resistance or disinclination to accept a new technology.
Conversely, Bala and Venkatesh's (2016) classification, which captures exploration and exploitation, among other benefit-seeking pursuits, provides a valuable lens suited to analysing the use of disruptive technology such as AI. By adding the two adaptive measures of ‘avoidance’ and ‘intention to revert’, Bala and Venkatesh demonstrated that employees may change or reproduce structures and avoid them or revert to previous, seemingly superseded ones.
Despite providing deep insights into the adaptation enacted by individuals and groups, these typologies only partially consider technology adaptation as part of business models suited to appropriate value. From the IS perspective, such typologies and conceptualization pertain to user–technology interactions rather than technology as an integral tool within a constellation of intra- and inter-organizational support mechanisms, as identified by Kung et al. (2015).
AST offers a useful theoretical lens for analysing the role played by HRM in technology adaptation (Turner, Morris and Atamenwan, 2019). The inter-relationships between AI adaptation and the strategic management of human resources within an organization (Shao, Li and Wang, 2022) justify our use of AST as a theoretical lens for this paper.
AI and SHRM
Several studies have explored the relationship between the use of technology in general (Apostolidis, Devine and Jabbar, 2022; Singh et al., 2021) – and of AI in particular (Chowdhury, Budhwar and Wood, 2024; Malik, Budhwar and Kazmi, 2023) – with HRM. In reference to the adaptation of AI applications by organizations, the impact on HRM may encompass several relevant areas, including job replacement, human–AI collaboration, recruitment, decision-making and training and development (Vrontis et al., 2022). In order for organizations to harness AI to drive the recruitment and retention of staff to foster innovation and productivity, SHRM and AI need to be aligned (Malik, Budhwar and Kazmi, 2023).
However, studies are increasingly focusing on the role HRM may play in developing, implementing and fostering innovations and technologies in business (Adisa et al., 2022; Amarakoon, Weerawardena and Verreynne, 2018). In so doing, SHRM can make incremental and/or radical changes in the structure and processes of an organization. Rapid technological developments, corporate strategies and the changing roles and expectations of employees require organizations to be agile and adaptive (Adisa et al., 2022; Apostolidis, Devine and Jabbar, 2022). Therefore, the literature highlights the strategic role that HRM can play in the adaptation of new technologies within an organization and the development of new business models and organizational structures.
Recent studies have emphasized how SHRM can foster the adoption of AI within an organization (Malik, Budhwar and Kazmi, 2023; Malik et al., 2022; Vrontis et al., 2022). However, such adoption does not necessarily warrant the continuous and smooth use of a technology, which calls for research to be conducted on technology adaptation, as stated in the previous section. We currently have only a tenuous understanding of the adaptation of AI in relation to SHRM. Our study is thus aimed at addressing this gap in the literature.
AI and value appropriation
Value is created through multiple interactions between different actors, including customers, technology, employees, processes and firms (Lusch and Vargo, 2014). Value appropriation denotes how any value created is appropriated for an organization and its internal and external stakeholders. Value appropriation is often viewed as preventing value slippage, highlighting the challenges involved in retaining value within an organization (Corsaro, 2020). The literature supports the notion that appropriate business models and industry architectures can help organizations achieve value appropriation.
This becomes more pertinent for business ecosystems in which independent and heterogeneous parties collectively generate value through iterative interactions (Thomas and Ritala, 2022). Although an emerging body of scholarly work (Albert-Cromarias, Asselineau and Blanchard, 2022; Volschenk, Ungerer and Smit, 2016) highlights the importance of studying the dynamic process of value appropriation, the literature is still focused mainly on outcome and output measures (Jell-Ojobor, Hajdini and Windsperger, 2022; Tower, Hewett and Saboo, 2021). According to Ellegaard, Medlin and Geersbro (2014), while value appropriation is commonly viewed as a result of resource utilization, it is crucial to highlight the dynamic process and interconnections among the actors involved in value appropriation. There is scholarly disagreement regarding the assessment and analysis of value appropriation. Consistent with Ellegaard, Medlin and Geersbro (2014), we consider it timely to contribute to this academic discourse by examining process-based value appropriation.
Research questions
Reflecting on the existing literature, we posited that a dedicated value appropriation strategy may improve organizational performance by efficiently deploying human capital (Lee and Miller, 1999). As such, HRM can contribute towards value appropriation within an organization by supporting the development of a competent, engaged and motivated workforce. Employee engagement and competencies are crucial for technology-led innovation and consequent value appropriation. New technologies like AI can drive value appropriation through various measures, including building capacity, streamlining operational processes and bringing product and process innovation (Mishra and Pani, 2020). Hence, an effective HR strategy can leverage AI adaptation to drive value appropriation, an aspect that has not been properly highlighted in the literature.
By conceptualizing value appropriation as a process, we can link value appropriation with organizational resources (AI) and functional strategies (e.g. HRM). We string these links together to develop the research question we address in our study: How and to what extent can SHRM shape AI adaptation and appropriate value within a dynamic business context?
Methodology
Research design and data collection
Our research objectives require a thorough analysis of various agents and processes and of their inter-relationships that shape AI adaptation and its outcomes. Technology adaptation is a function of constant interaction among human agents and organizational structures and systems, which iteratively influence each other (Dey et al., 2019). Accordingly, we identified the need for an exploratory study, underpinned by the interpretivist philosophy, to provide a nuanced understanding of employee experience, management strategy and organizations’ internal and external environmental dynamics.
We conducted semi-structured interviews with senior management personnel from 51 local/multinational, large and medium-sized organizations operating in a wide range of sectors in India. Our interview questions revolved around the organizational strategies followed in managing AI adaptation, HR strategies, the roles and responsibilities of different functional departments and the strategic approach to innovation and relevant challenges. Our interview guide is presented in Appendix A.
Data analysis
We recorded and transcribed all the interviews and entered them into NVivo 12 software for analysis. We conducted the data analysis using an inductive approach. Following the guidelines provided in the literature (Miles and Huberman, 1994), we established three themes (AI adaptation drivers and challenges, AI adaptation and value appropriation). We then divided each theme into multiple subthemes. Subsequently, two coders independently coded the transcripts in line with the themes and subthemes.
The two coders familiarized themselves with the interview transcripts to understand the context (Maxwell, 2008). They started with certain a priori codes and meticulously read through the transcripts for data-driven codes to identify emerging patterns. The first-order codes were contrasted against the a priori ones gathered from AST and relevant empirical works (e.g. Schmitz, Teng and Webb, 2016). For instance, following Schmitz, Teng and Webb (2016), exploration and exploitation were identified as two types of adaptation. Where there were no a priori codes, we referred back to the literature to find appropriate theoretically defined terminologies that captured the concepts. Once the two coders had reached a consensus, the codes were screened, revised and condensed. Through constant iteration and comparison between the codes and the relevant literature, third- and final-order constructs were determined. Figure 1 illustrates how the coding led to the taxonomy of AI adaptation.
Findings
Our findings enable us to depict how SHRM can shape different types of AI adaptation, leading to value appropriation within a dynamic business environment. We pay particular attention to the factors that constitute the value appropriation process. Accordingly, we identify and analyse the roles played by SHRM in determining AI-led value appropriation. We first identify the macro-, meso- and micro-factors linked with an organization's HR strategies and consequent AI adaptation. Subsequently, we create a taxonomy of AI adaptation based on the management's intent towards innovation and resulting value appropriation. Finally, by connecting the drivers, types of adaptation and HR roles, we define the value appropriation process within an organization.
Drivers and challenges of successful AI adaptation
For any organization, it is important to align a company's culture, structure and ways of working to support AI implementation. It is part of our HR policy to encourage creativity and innovation. (Tech-2)2
As a multinational conglomerate, we often do not have that level of flexibility, as we have to balance between local contextual issues and standard global strategy. (Cong-1)
The vendors and suppliers need to ensure that the transformation occurs properly. And, when we are implementing a new technology like Salesforce, our suppliers also need to have an alternative ready… Often, it is too difficult to develop AI in an isolated way when suppliers and distributors work in a more conventional system. (CG-3)
We are having great support from the government side, especially in India. And there are a lot of opportunities that they are giving to start-ups, which is enabling a lot of bigger organizations in the B2B context to do business. (SC-3)
As discussed in the following section, we use an intra-organizational dynamics perspective to examine the use of AI within a company.
AI adaptation strategies – exploitive, exploratory, expedient and emancipatory
We have observed that some of the organizations in our sample have used established AI tools to address specific challenges. These organizations have adhered to industry standards and practices and have viewed AI as a predetermined solution. This approach showcases an exploitation of AI. These organizations have primarily relied on external consultants to implement AI rather than investing heavily in the training and development of their internal staff.
The effective deployment of AI necessitates customization, which can be achieved through in-house expertise or the acquisition of specialized products. Nonetheless, tailoring AI applications can be a costly endeavour. Exploitative adaptation occurs when the absence of customization constrains organizations. They tend to sidestep the time-consuming process of experimenting with various AI applications to generate meaningful value.
It is important to note that regulatory frameworks can also limit exploitive adaptation. For instance, India has recently implemented a privacy law similar to the EU's General Data Protection Regulations. Nevertheless, this law is quite restrictive and results in increased costs and complexities for data mining and software companies (Govindarajan et al., 2019). A participant from a prominent Indian software company highlighted the challenges posed by governmental restrictions (refer to Table 1). Under this approach, the management aims to create value for the organization. However, its intentions are constrained due to financial, HR and operational limitations. Additionally, firms may face external environmental constraints. Consequently, management strives to achieve AI-led benefits by adhering to specific tasks. In such scenarios, the potential for value appropriation is restricted.
| Theme | Excerpts |
|---|---|
| Exploitive adaptation due to a lack of customization | We mostly use in-house products that we can try, explore and modify in our own way. I use some of the applications like AppDynamics. It analyses the data and comes up with a solution. But then you cannot customize it completely. (AM-1) |
| Exploitive adaptation due to regulatory restrictions | We have provision to store data in the cloud services based in Singapore, Mumbai and Abu Dhabi. But the new law in India restricts us to storing it only within India. (SC-1) |
| Exploitive adaptation due to financial and operational restrictions | Our product is built on a soft telephony layer, which is not great in India. In [the] United States it's very easy, because you can just use Skype anytime anywhere. But, in India, wide telephony is banned for entrepreneurial purposes. We try to get as much benefit as possible by building our product within these limitations. (SC-5) |
| Theme | Excerpts |
|---|---|
| Exploratory adaptation due to management and employee creativity | Chatbot is not always a solution. You may not get all insights with an ordinary chatbot. We tried a chatbot that didn't provide our desired outputs. Then, we tried the next level of chatbots, which is called cognitive insights chatbots. So, we explored different products and applications to get what we wanted. (DA-1) |
| Exploratory adaptation due to HR and financial support | Effective use of AI requires creativity. We offer special incentives to our employees who can come up with creative solutions. It is part of our HR policies. (TC-2) |
| Exploratory adaptation can be counterproductive for value appropriation | Trying out various options can be time and resource exhausting… Like, whenever a customer is on board, we would use AI, we would have some of the existing patterns learned from the past behaviour of existing customers. We have to try out different solutions and packages for different sets of customers. That's an expensive and time-consuming process. (FI-3) |
Several participants emphasized that their companies do not limit themselves to a specific use-case or application when it comes to their AI deployment strategy. Instead, they adopt an exploratory approach, experimenting with various forms of AI and use-cases to inform their AI deployment strategies. This is particularly apparent from an HRM perspective, as these organizations establish comprehensive training programmes and foster a culture of creativity among employees. They are open to experimentation, allowing employees the freedom to pursue innovative ideas that extend beyond their standard responsibilities.
Exploratory adaptation, however, requires financial backup and HRM support (see Table 2). As such, it can be resource- and time-consuming and requires staff incentives. Organizations should embed employee creativity within promotion and recognition schemes. It is imperative to provide training and R&D support for the exploration of AI applications. This ensures that users at all levels within an organization are well versed in the applications and can confidently recommend them based on their effectiveness.
An exploratory approach to adopting new technology may not always result in greater value due to the potential for trial and error and the overuse of organizational resources. As such, a lack of strategic vision – and a concern for the greater good of the organization and its other stakeholders – can lead to diminished value.
The external and internal constraints and value-diminishing nature of AI cause some organizations to avoid the complexity of its implementation by either reducing or limiting its use or avoiding it altogether. We term this adaptation strategy ‘expedient’. The word ‘expedient’ alludes to a ‘convenient’ approach, and it captures both the provision of ‘complete avoidance’ and ‘limited use’. Hence, the term is more inclusive compared to ‘avoidance’, as defined by Bala and Venkatesh (2016). As the planner and controller of e-commerce of a local conglomerate argued, the challenges linked to the use of AI can sometimes be insurmountable, prompting organizations to make little or no use of AI (see Table 3). Furthermore, a consistent and coherent strategy between HR and operations is imperative. Any newly introduced AI-driven application must align with employee expertise. Organizations may choose not to use AI, as training or replacing existing staff can be expensive.
| Theme | Excerpts |
|---|---|
| Expedient adaptation due to management scepticism | There is no 100% determined solution with AI capabilities. So, sometimes, the probabilistic nature of AI solutions doesn't go well with the end users because they don't understand the mathematical concepts behind it. So, that's the other challenge when it comes to the adaptation of AI by the company. (CG-3) |
| Expedient adaptation due to HR limitations and a lack of staff expertise | When I joined, I found that our accounting colleagues were only using Excel. They do not use high-end accounting software such as Tally or Zoho accounts. I introduced them to Tally and Zoho. They used it for a short period but decided to stick to Excel. This is not resistance – they adopted it but are not fully enjoying the tools… But, recruiting someone with those skillsets and experience would be expensive. Hence, we reverted to Excel. (MKT-2) |
We have identified another type of AI adaptation strategy (relevant excerpts are presented in Table 4). The participants who fell within this group raised issues pertaining to human emotions, ethics, governance and the democratization of AI. This reflects emancipatory motives behind using a technology, liberated from its original design and espoused with human emotions, ethics and social sustainability. According to AST, how a technology will be used is determined by the ‘spirit’ inscribed into it by its designers. The emancipatory nature of adaptation, underpinned by people-centred use, liberates the technology from its inscribed spirit and drives a contextually and ethically appropriate use. With the growing concerns regarding AI-enabled technology, such as ChatGPT, the ethically appropriate human-centred adaptation of AI has garnered importance both in India and worldwide. This innovative approach requires a strategic vision and care for the organization, its employees and other stakeholders.
| Theme | Excerpts |
|---|---|
| Emancipatory adaptation due to contextually appropriate management strategy | We found that drivers working in the rural areas are spiritual and they believe in astrology… However, we soon realized that the dates of birth on their official records are not correct. In India, the official dates of birth are often recorded during school exam registration. As astrology is linked with their dates of birth, we found that the vast majority of them have similar astrological records. Then we started to create separate information only on their star signs. This is not just e-business; this is nuanced understanding, related to people's emotions. (FI-4) |
| Emancipatory adaptation when ethics are embedded within organizational and HR strategies | I would say that, while all these strategies are getting developed, there is an ethical aspect to it. A lot of personal and organizational data are involved in our AI use. So, we are careful about our governance and data management. We have designed multiple layers of security system to ensure data protection. this wasn't there in the vendor's original product. We asked for it. (SC-2) |
| Emancipatory adaptation happens due to a bottom-up approach to organizational structure, management and HR practices | Democratization of a technology is the most important thing in the ability of the technology to be proliferated for use within [the] organization. AI provides a perfect scope for that, as it offers an open-ended use, underpinned by users’ ideas and experience. (IT-3) |
Value appropriation through AI adaptation
Our findings suggest that AI is not a panacea. However, the adaptation of AI within an organization can enable the effective use of resources. AI can be adapted to seamlessly connect employees across departments and leverage an organization's internal resources to meet the dynamic demands of the external environment. The four key facets of this value appropriation process can be identified as follows: (1) cross-departmental application and coordination; (2) employee capacity and efficiency; (3) process efficiency; and (4) product and service quality, as summarized in Table 5. Strategic HR can have direct or indirect implications for a number of these facets, including cross-departmental application and coordination, employee capacity and efficiency and process efficiency.
| Subtheme | Excerpts |
|---|---|
| Cross-departmental application and coordination | The information flow can be from other verticals as well, such as HR, Accounts, Logistics, Supply Chain and Finance. AI helps in getting a 360° view with all this information. Given the right algorithm and parameters, AI can help detect flaws in data that might be fraudulent in nature to begin with. (Cong-3) |
| Employee capacity and efficiency | Employees often consider AI as a threat; it will replace or reduce human involvement. We have debunked this assumption. We haven't opted for full automation or the complete replacement of human involvement. Rather, we have used AI to complement employee expertise, while simultaneously increasing staff capacities to deliver in a prompt and effective manner. (FI-5) |
| Process efficiency | Due to AI use, we get a much clearer picture on what is happening in the market, what is the basket size for an outlet, which SKUs are moving and which are not. So now our data are much more sanitized and are at a real-time level. Accordingly, our production process has become more efficient. (CG-4) |
| Product and service quality | Through the use of AI, we can now create customized products for our customers. This is where we are offering value to our customers by leveraging and optimizing our internal capabilities. (IT-4) |
A taxonomy of AI adaptation
We placed the four categories of adaptation within a 2 × 2 matrix (see Table 6). The vertical dimension represents the ‘extent of managerial intent for innovation’ – which drives adaptation – and the horizontal dimension represents ‘managerial desire for value appropriation’.
| Extent of value appropriation (higher) | Extent of value appropriation (lower) | |
|---|---|---|
| Management intent for innovation (more favourable) | Emancipatory (emancipate to maximize) | Exploratory (explore to satisfice) |
| Management intent for innovation (less favourable) | Expedient (escape to maximize) | Exploitive (exploit to satisfice) |
The emancipatory approach, one of the novel contributions of this research, is driven by a proactive and favourable management that aims to maximize value for all stakeholders and take a strategic approach to value appropriation through the suitable value-laden use of AI. An organization's willingness and ability to engage in value appropriation are demonstrated by increased process efficiency, product quality and an ethical and people-centred adaptation of AI. This occurs when value appropriation is maximized through the concurrent intent and effort of the management to adapt AI in an innovative, responsible and contextually appropriate manner.
Expedient adaptation, on the other hand, may achieve value appropriation through production, process and employee efficiency. However, the management is reluctant to innovate and may choose to avoid using AI. As mentioned above, exploratory adaptation may involve customization and trial-and-error approaches that may not lead to the efficient management of production, processes and employees. However, such endeavour is led by a managerial intent to innovate. By avoiding the trial, experimentation and introduction of AI tools, organizations seek to reduce costs and increase value in a less innovative manner.
Finally, exploitive adaptation (exploitation to satisfice) occurs when the managerial intent to innovate is limited and involves lower value appropriation than an expedient approach. Exploitive adaptation involves incremental changes underpinned by a conservative approach.
Hence, the matrix provides a comparative scenario for AI adaptation considering innovation and value appropriation. Emancipatory and expedient adaptation exhibit higher value appropriation than their exploratory and exploitive adaptation counterparts, respectively. On the other hand, emancipatory and exploratory appropriation result from managerial intent that is more innovative than that involved in either expedient or exploitive appropriation.
Discussion and conclusion
Organizations strive to create value through their business model, process and/or product innovation. However, creating value does not necessarily lead to sustained benefits for an organization or its various stakeholders, including its own employees (Corsaro, 2020). Hence, value appropriation is considered to assess the extent to which an organization retains the value created. In our study, we analysed the role played by SHRM in harnessing value appropriation by driving AI adaptation.
By analysing the findings through the lens of AST, we found that the nature and mode of the interaction between human and institutional agents, along with organizational structures, practices, rules and macro-environmental dynamics, can be identified as both facilitators and inhibitors of AI adaptation.
Theoretical contributions
This paper accomplishes its objective by presenting empirical evidence that leads to significant theoretical contributions. The primary contribution involves the creation of a model of value appropriation supported by SHRM. Additionally, our findings introduce a novel AI adaptation matrix, which offers insights into achieving various AI adaptation strategies aligned with management's intention to innovate and to appropriate value. The integration of these contributions establishes a solid conceptual framework for understanding how SHRM can effectively leverage different AI adaptations to appropriate value.
The role of SHRM in AI-led value appropriation
By exploring value appropriation as a function of managerial innovation and technology adaptation, we complemented the positivist perspective of value appropriation that dominates the extant literature (Mizik and Jacobson, 2003; Ritala and Tidström, 2014). Our perspective is inspired by Yang et al. (2021), who argue that value cannot be appropriated without the collaborative support of the drivers within an organization. The complementarity and mobility of the components within a business model are necessary for innovation, which subsequently enables the creation and appropriation of value. Our research has revealed the interconnected nature of these triangular relationships, as illustrated in Figure 2. Recognizing the diverse approaches of management towards innovation and SHRM, we have delineated different types of AI adaptations that can be applied within a specific organizational setting to achieve value appropriation.
Our findings demonstrate that SHRM is pivotal in determining how AI-led value is created and appropriated within an organization. This is especially applicable to AI, which requires upskilled and boundary-spanning roles due to the importance of cross-departmental coordination and application in value appropriation. Our findings represent a response to the call in the literature (Morris et al., 2017) to consider employee roles and intra-organizational dynamics in value appropriation. Our findings show that organizational strategy defines SHRM measures and the extent of business model innovation, which jointly constitute AI-led value appropriation. AI implementation may require: (i) training and support from HR; (ii) data input from the operations, logistics and sales teams; and (iii) analysis and strategic decision-making from senior management. As such, we present a conceptual model (Figure 2) that summarizes how SHRM and organizational intent for innovation shape the technology adaptation which leads to value appropriation.
The AI adaptation matrix
As per Bala and Venkatesh (2016), many of our participants are inclined towards engaging in exploration and exploitation to realize the benefits of AI. Exploitation occurs when there is a lack of managerial intent to explore and innovate. On the other hand, exploration occurs due to innovative management practices. However, in both cases, an organization does not necessarily have appropriate value. Hence, we depart from Bala and Venkatesh (2016) by defining these two strategies as value-satisficing rather than value-maximizing means.
We consider both exploration and exploitation part of faithful appropriation, described as part of AST (DeSanctis and Poole, 1994). In both cases, adaptation is conducted following the spirit of a technological application. The novelty of our contribution is defined by introducing the two additional categories of ‘expedient’ and ‘emancipatory’ adaptation. Based on AST, the unfaithful appropriation of technology occurs when the agents ignore the spirit of the design and create a new technology use structure. We argue that unfaithful appropriation can take place in two ways: either by discontinuing the use altogether or by creatively adapting technology within contextual and ethical boundaries. We termed the first category ‘expedient’ and the second ‘emancipatory’.
Bala and Venkatesh (2016) introduced the concepts of ‘avoidance’ and ‘reversion’ as two additional adaptation strategies alongside ‘exploitation’ and ‘exploration’. We have amalgamated avoidance and reversion into a new category called ‘expedient’ adaptation. From an organizational standpoint, it is beneficial for SHRM to perceive avoidance and reversion as a unified type. Both avoidance of new and reversion to old technology account for the discontinuation of technology use, as denoted by our ‘expedient’ classification. We amalgamate the two from the perspective of maximizing value. Whether or not managers intentionally avoid or revert, they may harbour the same strategic intent, encapsulated within our framework by the construct of ‘expedient’ adaptation.
In addition, we have introduced a new form of adaptation called ‘emancipatory’ that has not been addressed in previous studies. Emancipatory adaptation occurs when technology goes beyond the intentions of its designer, as noted in the findings. Existing IS literature refers to emancipatory design, which aims to incorporate ethical standards in IS design while ensuring value maximization, as proposed by Hirschheim and Klein (1994). While Hirschheim and Klein (1994) considered emancipation design feasible within the existing power and organizational structure, we argue that the emancipatory approach mandates structural adjustments in the context of technology adaptation. This is essential to ensure AI's ethical and contextual use and adaptation, which generate an appropriate value for an organization. Emancipatory adaptation is particularly important for AI due to its inherent characteristics (refer to the summary in Appendix D), which distinguish it from traditional IT applications and integrate it within an organizational and SHRM context.
Implications for practice and policy
In January 2024, amidst the fear that their products would not be able to provide the expected benefits, top AI-related companies lost $190 billion in the stock market.3 As concerns for similar scenarios in the future remain, for example, due to conflicting issues concerning data privacy regulations (Dorotic, Stagno and Warlop, 2024), many companies are likely to take a conservative approach to AI adaptation. Expedient adaptation suits management with a conservative approach to innovation and a focus on minimizing costs. Organizations facing financial constraints and limited resources will likely benefit from a more cautious or limited use of AI. In this instance, SHRM has to work closely with top management to discontinue using AI tools not aligned with employee skills and experience.
Nevertheless, organizations with a conservative approach may opt for appropriate value through exploitative AI adaptation. They aim to minimize data sourcing and storage costs, reduce product customization and improve employee efficiency. To achieve that, SHRM must prioritize enhancing internal capabilities over seeking external expertise. Additionally, SHRM may centralize AI-related functions as a measure to save costs and appropriate value.
Embracing explorative adaptation necessitates an innovative approach. Nevertheless, management might struggle to maximize value appropriation because of the substantial expenses associated with implementing state-of-the-art, tailor-made AI tools, training and hiring competent personnel. This approach could suit a well-endowed organization operating in a fiercely competitive industry. In light of the swiftly evolving AI landscape, SHRM should prioritize enhancing employee skills and adaptability to keep abreast of these changes.
In response to increasing ethical issues and concerns about privacy, innovative SHRM should also consider a more proactive approach to addressing these issues, by adopting an emancipatory adaptation. This approach aims to ensure compliance with ethical norms and regulations while maximizing value through innovation. SHRM should allocate resources to develop employees and foster their creativity beyond the functional applications of AI, focusing on upholding ethical and legal standards. Therefore, SHRM should promote cross-departmental collaboration, encourage boundary-spanning and intrapreneurial management practices and engage external experts when necessary and develop internal staff skills.
Limitations and future research directions
-
Within a more innovative corporate strategy, SHRM will resort to exploratory AI adaptation to satisfice value.
-
Within a less innovative corporate strategy, SHRM will resort to exploitive AI adaptation to satisfice value.
-
Within a more innovative corporate strategy, SHRM will resort to emancipatory AI adaptation to maximize value.
-
Within a less innovative corporate strategy, SHRM will resort to expedient AI adaptation to maximize value.
Future research can also develop specific scales for emancipatory and expedient adaptation, while studies could differentiate between organizations according to their size and the extent of their internationalization to ascertain whether these factors exert an influence on adaptation and value appropriation methods. Furthermore, by conducting research in a multinational context, rather than selecting a single country, future researchers could examine the cultural, technological and socioeconomic factors affecting AI usage in multiple locations, and the different opportunities and challenges encountered by organizations in AI adaptation.
Biographies
Sanjit K. Roy is Professor of Marketing and Service Science at Edith Cowan University in Perth. He is an Associate Editor for the European Journal of Marketing. His research interests include the impact of technology on marketing and transformative service research. He has published in journals including the British Journal of Management, European Journal of Marketing, Journal of Business Research, Internet Research, Information Systems Frontiers and Journal of Services Marketing.
Bidit Lal Dey is Associate Professor of Marketing and Director of Accreditation at Sheffield University Management School, University of Sheffield. His work demonstrates thought leadership in consumer behaviour and digital and international marketing. He has published extensively in leading management, marketing and information systems journals, including the British Journal of Management, Journal of the Association for Information Systems, Journal of World Business, Journal of Business Research and International Marketing Review.
David M. Brown is Associate Professor of Marketing at Heriot-Watt University, Edinburgh. An established textbook author, David's research interests centre around organizations’ social and environmental sustainability initiatives, and particularly the sustainable co-creation of value. His major focus currently is on how pubs, brewers, distillers and their consumers can evolve towards a sustainable future. He also publishes on internal marketing, serious leisure tourism and pedagogical inclusivity, and is Co-Chair of the Academy of Marketing's Services SIG.
Aman Abid is Lecturer in Marketing at the UNE Business School, University of New England, Australia. He attained his PhD in Marketing from The University of Western Australia in 2022. His research explores the role of technologies in the field of marketing. His research has appeared in various marketing journals, including the Journal of Strategic Marketing, Journal of Marketing Management, Marketing Intelligence and Planning and Australasian Marketing Journal.
Chrysostomos Apostolidis is Associate Professor of Marketing at Durham University Business School and Co-Director of the Centre for Consumers and Sustainable Consumption. He is also an Associate Editor of the Journal of Services Marketing. His research interests lie at the intersection of digital technologies, sustainability and marketing. He has published in leading journals, including the British Journal of Management, European Journal of Operational Research, Journal of World Business, European Journal of Marketing, Technological Forecasting and Social Change and Journal of Business Research.
Michael Christofi is affiliated with Cyprus University of Technology and University of Vilnius. His research interests lie on the interface of innovation, marketing and entrepreneurship. His research has been published in premier outlets, such as the Journal of Product Innovation Management, British Journal of Management and Journal of World Business. Dr Christofi serves as a Senior Editor for Information Technology & People, an Associate Editor for International Marketing Review and a guest editor for leading journals, including the British Journal of Management, Journal of Business Ethics and Technovation.
Shlomo Tarba is Chair (Full Professor) of Strategy & International Business, University of Birmingham. He is a Fellow of the Academy of Social Sciences. His research interests include M&A, agility and AI. His papers are published in the Journal of Management, Journal of Product Innovation Management, Journal of Organizational Behaviour, Human Relations, Academy of Management Perspectives and Journal of Service Research.
References
- 1 Top 3 challenges in AI faced by UK financial services firms – AI Business.
- 2 Refer to Appendix B for more details of the participants.
- 3 AI companies lose $190 billion in market cap after Alphabet and Microsoft report | Reuters.
