Embedding Benevolence into Business Processes - AI-Enabled Routinized Benevolence

Abstract

Embedding benevolent practices into day-to-day business operations can lead to sustained success without compromising profitability. However, many of the benevolent initiatives remain isolated and are not systematically integrated into routine business operations, limiting their broader impact. Addressing this gap, this study introduces the concept of Routinized Benevolence (RB), the integration of benevolent actions into routine business operations aimed at enhancing customer well-being. The study aims to deepen the conceptual understanding of RB and examine how for-profit organizations can embed benevolent practices into their business processes, particularly using AI as an enabler. Employing a Design Science Research approach, this study develops a few interconnected artifacts. The first phase explores what characterizes RB by introducing the RB Canvas, which outlines its foundational elements. By examining patterns of RB through the canvas, the study develops a set of distinct RB categories. The second phase investigates how organizations can embed benevolence into routine operations and how AI can be used as an enabler. This involves the development of procedural models specific to each RB category, further extended through the lens of AI enablement. Together, these artifacts constitute a cohesive framework that provides both conceptual insight and actionable guidelines to integrate benevolence into routine business operations while sustaining economic success.

Keywords

• eol>Routinized Benevolence
• Canvas
• Categorization
• Procedural Models
• AI-Enablement

1. Introduction

The traditional notion of prioritizing shareholders has been challenged during the past decade, prompting organizations to adopt a more holistic approach that balances economic objectives with pro-social values [ 1 ]. Rather than viewing pro-social practices as a trade-of against profitability, organizations can achieve sustained success by strategically integrating these initiatives into their core operations [ 2 ]. However, these practices often remain isolated and confined to higher-level strategic statements such as corporate mission statements or sustainability reports, rather than being embedded into routine practices that align with business objectives and drive long-term success.

As highlighted by Kramer and Porter [ 3 ], the shift toward the creation of shared value is reshaping corporate thinking, allowing companies to generate societal value alongside economic gains. However, despite the growing emphasis on shared value, traditional business processes remain predominantly focused on economic eficiency and are at risk of becoming obsolete amid increasing demands for transparency, fairness, and competitive adaptability [ 4 ]. While the concept of shared value ofers a strong vision for enhancing societal well-being, BPM research still lacks actionable frameworks for embedding benevolence into routine operations, particularly in ways that leverage emerging technologies.

"Benevolence" embodies the idea of doing good. It is defined as showing genuine consideration and sensitivity towards others [ 5 ]. Although benevolence has been extensively studied in psychological and not-for-profit organizational contexts, its application within the for-profit sector remains largely underexplored. This may be due to the limited adoption of benevolent practices in corporate environments, as organizations often grapple with the paradoxes between investing in benevolence (aimed at securing long-term gains) and pursuing short-term profitability [ 3 ].

Rosemann et al., [ 6 ], argue that organizations can sustain economic performance by integrating benevolent practices into their day-to-day operations. Building on this foundation, this study presents the concept of Routinized Benevolence (RB), the voluntary, mutualistic actions embedded into routine operations of for-profit organizations that prioritize the well-being of customers. Despite its potential, RB has received limited attention in both academic and practical domains, leaving its conceptual foundations underdeveloped. There is a lack of clarity on the defining characteristics of RB, its distinction from related constructs, the mechanisms through which it can be embedded in diverse for-profit organizational contexts, and the role emerging technologies play in enabling and scaling such practices.

Thus, a necessary first step in this endeavor is to conceptualize RB by examining its core elements including its meta characteristics, triggers, impacts, the operational patterns and associated risks. To meaningfully interpret and compare these patterns, and to inform theory building, it becomes necessary to identify the categories of RB, focusing on organizing diverse RB initiatives into coherent groups based on shared patterns of actions. A key dimension of this inquiry involves exploring how forprofit organizations can integrate benevolence into routine operations in a scalable, proactive, and timely manner. This involves challenges such as identifying the right moments to act, personalizing responses, and ensuring consistency. These challenges highlight the need for capabilities like pattern recognition, real-time decision-making, and personalization, which go beyond traditional rule-based systems. In this context, AI stands out as a powerful enabler by providing data-driven insights and supporting sustainable business practices [ 7 ]. Furthermore, although AI is increasingly applied to enhance productivity, there is a growing call for embedding AI in more human-centric processes [ 8 ], confirming its relevance to the operationalization of RB.

Consequently, this study aims to address three main research objectives (RO) as outlined below. RO1: To identify and define the key characteristics of Routinized Benevolence (RB)

RO2: To develop practical guidance on how for-profit organizations can integrate benevolence into their routine operations

RO3: To explore how AI can be leveraged to enable and scale the implementation of RB Employing a Design Science Research (DSR) approach, this study develops a set of interconnected artefacts as a solution to a real-world problem. To address the first research objective, the study investigates the phenomenon of RB by examining what RB entails, how it manifests. It begins with the formulation of a formal definition of RB, derived from a literature analysis. This is followed by the development of a conceptual model, presented in the form of a canvas that captures the key dimensions of RB, informed by an inductive analysis of case examples form the practise. RB is introduced as a novel construct that extends the concept of value creation within the BPM domain. By examining the empirical cases mapped onto this canvas, the study identifies recurring patterns and contextual distinctions in how RB is enacted across diverse industries, facilitating a categorization of RB. At this stage, the study has produced the initial version of RB Canvas as a foundational artefact. In the subsequent phases, a categorization of RB will be developed based on the patterns identified through case analysis. To address the RO2, procedural models will be designed for each identified RB category, illustrating how benevolence can be systematically embedded into routines. These models will then be extended to address the RO3 by identifying how AI can serve as an enabler in supporting or scaling RB.

2. Related Work

2.1. Routinized Benevolence

Benevolence, in general, was defined as the act of preserving and improving the well-being of individuals with whom one has frequent personal interactions [ 9 ]. Although it was extensively studied in psychological and non-profit contexts, only a few researchers [ 10, 11 ] have studied benevolence in the for-profit organizational context. Organizational benevolence was defined as the commitment of a company to advance the well-being of an external stakeholder group as a core objective, reflected in a behavioral inclination where the primary goal is to serve the interests of others above self-interest [ 10 ], recognizing it as an altruistic approach, achieved without having any egocentric motive. In contrast, we argue that for-profit organizations often view benevolence not as the final objective but as a strategy that ultimately supports profit generation. Rosemann et al. [ 6 ] similarly contend that benevolence can enhance customer value, build trust and loyalty, and lead to sustained economic benefits within for-profit settings. They further highlighted the importance of examining the role of technology in enabling RB, a goal that this paper aims to address.

Benevolence becomes genuinely embedded when it is an integral component of routine operations, seamlessly integrated into everyday business practices, rather than treated as an incidental or auxiliary efort [ 11 ]. As emphasized by Sele et al., [ 12 ], routines are capable of addressing grand challenges, as they consist of recurring patterns of action that can be intentionally designed to generate lasting impact — reinforcing the importance of embedding benevolence into the very fabric of business processes. Using existing literature on benevolence and insights from illustrative case examples from practice, this study defines RB as “the voluntary acts embedded into routine business operations of for-profit organizations, which prioritize the well-being of customers, and undertaken as an intermediate goal, with the expectation of long-term gains”. This definition rests on four key criteria: (1) It must prioritize the well-being of customers, (2) RB should be embedded into routine business operations, rather than executed as isolated or one-of charitable acts. (3) It should reflect a mutualistic approach, where the organization incurs a cost or forgoes immediate gains with the expectation of long-term benefits. (4) RB must be voluntary in nature, not necessarily driven by laws, regulations, or external reporting requirements.

RB difers from related concepts such as Corporate Social Responsibility (CSR), shared value, corporate philanthropy, and compassion in several important ways. CSR, while socially driven, is often shaped by external regulations and tends to be implemented as discrete initiatives for visibility or compliance [ 13 ]. In contrast, RB is not shaped by regulations or external mandates, is embedded into a firm’s regular operations, and specifically targets customers. Shared value similarly seeks to address societal issues while benefiting the company [ 14 ], but its broader societal orientation contrasts with the focus of RB on customer well-being. Unlike corporate compassion and philanthropy, which are altruistic [ 15 ], RB aims to create mutual value, benefiting both customers and the organization in the long term.

2.2. AI-Enabled Routinized Benevolence

Enabling RB with technology requires advanced analytical techniques and real-time data processing, allowing organizations to anticipate customer needs, behaviors, and trends, thereby facilitating more informed decision making. AI plays a vital role in enabling businesses to optimize their operations by forecasting customer behaviors and ofering personalized experiences [ 7 ]. While AI for social good has been explored in research [ 16, 17 ], real-world applications that integrate AI into benevolent actions remain scarce, likely due to the uncertainty of long-term value and practical implementation. In general, AI was defined as the ability of a machine to exhibit human intelligence, such as perceiving, reasoning, learning, and interacting [ 16 ]. Additional capabilities encompass: problem solving, learning, knowledge representation, communicating, natural language processing and acting [ 18 ]. Building on this and related studies [ 19 ], we define AI-enablement as the presence of at least one of these capabilities.

The case example analyzed in this study (refer Section 3) highlights the role of AI or the potential of using AI to enable RB. For example, Tesla uses AI to identify vehicles that require extended battery capacity during emergencies. These cases reveal how AI enhances RB by identifying opportunities for automation, augmentation, or autonomous execution, allowing organizations to embed benevolence into their core operations in a scalable way. In this context, automation refers to rule-based systems that replace human labor; augmentation involves data-driven insights that assist human decision-making; and autonomous execution denotes AI systems capable of acting independently [ 20 ].

3. Research Methodology

The Design Science Research (DSR) approach was chosen to address the research questions by developing practical, actionable solutions to a real-world problem. Aligned with the three research objectives (see Section 1), the study is designed to produce three key conceptual artifacts: (i) the RB Canvas and Categorization of RB (addressing RO1), (ii) procedural models for operationalizing RB (addressing RO2), and (iii) AI-enabled procedural models for operationalizing RB (addressing RO3).

Pefers et al. [ 21 ] proposed the Design Science Research Process (DSRP) model, which serves as both a nominal process model to conduct DSR and a mental model to present DSR in the field of Information Systems. It consists of six main stages: problem identification and motivation, defining the requirements of the solution, design and development of the artifact, demonstration, evaluation, and communication of outcomes. Aligned with these six stages, Figure 01 illustrates the DSR approach adopted in this study. The elements highlighted in green represent the current status and the progress made to date.

The research begins with the problem identification phase, which involved conducting a preliminary literature review to identify gaps in existing knowledge around embedding benevolence into business processes. Following the problematization approach suggested by Alvesson and Sandberg [ 22 ], this study questions the prevailing view that benevolence in business is inherently altruistic, exceptional and peripheral. This critical lens paves the way for reconceptualizing benevolence as a practice that could be embedded into organizational routines while sustaining profitability. This academic inquiry was enriched by discussions with practitioners and colleagues. Based on the identified problem, the requirements for the first artifact, the RB canvas, are determined as follows.

R1: The RB canvas should identify key facets of RB - Since this study seeks to conceptualize RB, the artifact should capture and demonstrate its key facets including the underlying drivers of RB, its core characteristics, associated risks, and the impacts it generates within organizational contexts.

R2: The RB canvas should capture the patterns of actions involved in operationalization of RB across diferent corporate settings.

R3: The RB canvas should enable the identification of distinct categories of RB - Based on the nuances in the patterns of actions, the artifact should identify distinct categories of RB.

This study has developed an initial version of the RB canvas, grounded in literature and a real-world case example analysis. The illustrative cases were drawn from both the existing literature on benevolence and publicly available resources such as business news sites (e.g., Forbes, McKinsey Insights), company websites and social networks where customers share their experience (e.g., Reddit, LinkedIn). These cases were evaluated against the predefined RB criteria, (see Appendix - Part A). Upon identifying an example, we gathered additional details from industry reports, corporate documents, and social networks to build a case repository. Ten qualified cases were identified; five were selected to develop the artifact based on data richness and industry diversity, while the remaining five are reserved for later demonstration and evaluation (see Appendix Part B).

The case analysis consists of a three-cycle coding approach [ 22 ]. In the first cycle of coding, we looked at the cases line by line to extrapolate the ways in how they describe the benevolent act, with a particular focus on its drivers, operational design, characteristics, impacts and risks. Then, in round two, the research team collaboratively refined and reorganized the initial codes, grouping them into higher-level themes based on commonalities and relationships between diferent aspects of benevolent acts. In round three, several workshops were conducted with the research team to analyze and interpret the identified themes. To synthesize the identified themes, the canvas approach emerged as a viable framework for demonstrating the results, inspired by the initial Business Model Canvas (BMC).

A closer examination of prior studies [ 23, 24 ] underscores the versatility of the BMC as an efective tool for visualizing interrelationships among organizational components, thus supporting strategic decision-making. Furthermore, the themes inductively derived from the case analysis demonstrated strong alignment with the structure of the BMC framework. This convergence ofers a compelling rationale for adapting the BMC as a foundational model to represent the key elements of RB. The RB canvas along with the definitions of its key elements are available in part C of the appendix.

The demonstration phase is carried out by mapping five real-world case examples onto the RB Canvas (see Appendix – Part D), assessing how efectively the canvas delivers its intended outcomes. To evaluate the validity of artifacts, the study will conduct case studies with practitioners to assess their practical value and applicability [ 23 ]. The planned future work is outlined in the following section.

4. Discussion

This study contributes to the BPM discipline by conceptualizing RB - a notion that has received limited attention in academia and practice. Unlike traditional views that treat benevolence as discretionary acts, RB is positioned as a processual, repeatable approach embedded into routine operations. Through an inductive case analysis, this study has developed a canvas-based framework that captures the key elements of RB. This framework extends BPM body of knowledge by incorporating humanistic and stakeholder-centered objectives. It also aligns with recent BPM calls, particularly Value-Driven BPM [ 24 ] by embedding ethically grounded, customer-focused practices into business processes.

Furthermore, by exploring how AI can enable and scale RB,this study opens a new line of inquiry that connects emerging technologies with human-centric process design. This posits AI not just as a tool for eficiency, but as a catalyst for ethical and socially responsive BPM practices. Together, the integrated artifacts develop in this study provide a holistic foundation for embedding benevolence into organizational routines. For practitioners, the study provides actionable guidelines and design considerations for integrating emerging technologies into customer-facing processes in a responsible and value-driven manner. As part of the planned future work, this study will develop a categorization of RB, based on patterns identified through case-based canvases. Building on this and drawing on the literature on procedural models, it will construct procedural models to guide the systematic implementation of RB, further enhanced through the lens of AI enablement. Collectively, these artifacts, the RB canvas, categorization, and procedural models, will be empirically evaluated through case studies and validated through workshops with industry practitioners. This phase aims to ensure both the practical relevance and operational feasibility of implementing RB. In addition, a survey with workshop participants will be conducted to further validate the artifacts in terms of usability, usefulness, and acceptance.

Through this consortium, I seek constructive feedback on assessing the efectiveness of the proposed artifacts and advancing the theoretical contribution to the emerging discourse on Responsible BPM.

5. Appendices

The appendix is available via https://tinyurl.com/y3cjxj5d

Declaration on Generative AI

The author(s) have not employed any Generative AI tools.

References

[1] M.-C. Boudreau , Corporate social responsibility in the it industry: A stakeholder perspective ( 2023 ).

[2] R. G. Eccles , I. Ioannou, G. Serafeim, The impact of corporate sustainability on organizational processes and performance , Management science 60 ( 2014 ) 2835 - 2857 .

[3] M. R. Kramer , M. Porter , Creating shared value , volume 17 , FSG Boston, MA, 2011 .

[4] D. Peppers , M. Rogers , Extreme trust: The competitive necessity of proactive trustworthiness , in: Marketing Wisdom , Springer, 2018 , pp. 1 - 16 .

[5] K. Atuahene-Gima , H. Li , When does trust matter? antecedents and contingent efects of supervisee trust on performance in selling new products in china and the united states , Journal of Marketing 66 ( 2002 ) 61 - 81 .

[6] M. Rosemann , N. Ostern , M. Voss , W. Bandara, Benevolent business processes-design guidelines beyond transactional value , in: International Conference on Business Process Management , Springer, 2023 , pp. 447 - 464 .

[7] D. Van Chau , J. He , Machine learning innovations for proactive customer behavior prediction: A strategic tool for dynamic market adaptation , 2024 .

[8] S. Thiebes , S. Lins , A. Sunyaev , Trustworthy artificial intelligence, Electronic Markets 31 ( 2021 ) 447 - 464 .

[9] S. H. Schwartz , A. Bardi , Value hierarchies across cultures: Taking a similarities perspective , Journal of cross-cultural Psychology 32 ( 2001 ) 268 - 290 .

[10] A. J. Beveridge , M. A. Hollerer , Theorizing organizational benevolence, Organization Science 34 ( 2023 ) 1864 - 1886 .

[11] M. Rosemann , W. Bandara , N. Ostern , M. Voss , The benevolent enterprise: How to operationalise and scale doing good ( 2023 ).

[12] K. Sele , C. A. Mahringer , A. Danner-Schröder , T. Grisold , B. Renzl , We are all pattern makers! how a flat ontology connects organizational routines and grand challenges , Strategic Organization 22 ( 2024 ) 530 - 549 .

[13] R. G. Boutilier , A measure of the social license to operate for infrastructure and extractive projects , Available at SSRN 3204005 ( 2017 ).

[14] P. S. Menghwar , A. Daood , Creating shared value: A systematic review, synthesis and integrative perspective , International Journal of Management Reviews 23 ( 2021 ) 466 - 485 .

[15] J.-P. Vergne , G. Wernicke, S. Brenner , Signal incongruence and its consequences: A study of media disapproval and ceo overcompensation , Organization Science 29 ( 2018 ) 796 - 817 .

[16] J. Cowls , A. Tsamados , M. Taddeo , L. Floridi , A definition, benchmark and database of ai for social good initiatives , Nature Machine Intelligence 3 ( 2021 ) 111 - 115 .

[17] N. Tomašev , J. Cornebise , F. Hutter , S. Mohamed , A. Picciariello , B. Connelly , D. C. Belgrave , D. Ezer , F. C. v. d. Haert, F. Mugisha , et al., Ai for social good: unlocking the opportunity for positive impact , Nature Communications 11 ( 2020 ) 2468 .

[18] S. J. Russell , P. Norvig , Artificial intelligence: a modern approach , pearson, 2016 .

[19] C. Rzepka , B. Berger, User interaction with ai-enabled systems: A systematic review of is research ( 2018 ).

[20] S. Raisch , S. Krakowski , Artificial intelligence and management: The automation-augmentation paradox , Academy of management review 46 ( 2021 ) 192 - 210 .

[21] K. Pefers , T. Tuunanen , C. E. Gengler , M. Rossi , W. Hui , V. Virtanen , J. Bragge, The design science research process: A model for producing and presenting information system research , in: 1st International Conference on Design Science in Information Systems and Technology (DESRIST) , volume 24 , 2006 , pp. 83 - 106 .

[22] J. Saldaña , M. Omasta , Qualitative research: Analyzing life, Sage Publications , 2016 .

[23] K. M. Eisenhardt , Building theories from case study research , Academy of management review 14 ( 1989 ) 532 - 550 .

[24] J. vom Brocke , C. Sonnenberg, Value-orientation in business process management , Springer, 2014 .