A Complex Leadership Perspective on Generative AI Adoption in SMEs: The Interplay of TAM, TMT, and RBV PDF Free Download
1 / 23/23
100%
Received: 7 October 2025
Revised: 1 December 2025
Accepted: 8 December 2025
Published: 16 December 2025
Citation: Peñarroya-Farell, M., Vaziri,
M., Soto Rivera, S. K., & Miralles, F.
(2025). A Complex Leadership
Perspective on Generative AI
Adoption in SMEs: The Interplay of
TAM, TMT, and RBV. Administrative
Sciences,15(12), 494. https://
doi.org/10.3390/admsci15120494
Copyright: © 2025 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license
(https://creativecommons.org/
licenses/by/4.0/).
Article
A Complex Leadership Perspective on Generative AI Adoption
in SMEs: The Interplay of TAM, TMT, and RBV
Montserrat Peñarroya-Farell * , Maryam Vaziri , Sasha Katalina Soto Rivera and Francesc Miralles
La Salle’s Smart Society Research Group, Ramon Llull University, 08022 Barcelona, Spain;
maryam.vaziri@salle.url.edu (M.V.); sashakatalina.soto@salle.url.edu (S.K.S.R.);
francesc.miralles@salle.url.edu (F.M.)
*Correspondence: montserrat.penarroya@salle.url.edu
Abstract
Although Generative Artificial Intelligence (GenAI) is one of the strategic choices for digi-
tal transformation in small and medium-sized enterprises (SMEs), its adoption remains
constrained by leadership decision-making that must balance strategic aspirations with
resource limitations and organizational inertia. Organizational leadership must face the dy-
namic and complex characteristics of digital transformation in the knowledge era. Drawing
on Complexity Theory and integrating the Technology Acceptance Model (TAM), Tempo-
ral Motivation Theory (TMT), and the Resource-Based View (RBV), this study proposes
a conceptual framework reflecting distinct strategic leadership orientations. Following
a qualitative approach based on semi-structured interviews with SME leaders and an
Interpretative Phenomenological Analysis (IPA) this conceptual framework contributes
by reframing GenAI adoption as a complex, nonlinear process rather than a straightfor-
ward diffusion model, that includes four strategic profiles (Strategic Adopters, Aspiring
Adopters, Opportunistic Adopters, and Operational Stabilizers) that affect a dynamic re-
lationship between three key adoption dimensions: intention, motivation, and resource
allocation. SME leaders can benefit from a delimitation of their strategic and operational
goals while overcoming adoption barriers.
Keywords: GenAI; SMEs; leadership behavior; technology adoption; strategic alignment;
complexity theory
1. Introduction
In today’s rapidly evolving technological landscape, Small and Medium Enterprises
(SMEs) face unprecedented opportunities and challenges. To build resilience and foster
digital transformation, key factors such as dynamic capabilities, digital inclusion, lead-
ership orientation, and knowledge management play a pivotal role (Sagala & ˝
Ori,2024).
Among emerging technologies, Generative Artificial Intelligence (GenAI) stands out for its
potential to enhance innovation, productivity, and competitiveness (Dwivedi et al.,2024).
Unlike traditional AI, which analyzes existing data, GenAI creates new content in various
formats—including text, images, video, or music—based on its training data (Goodfel-
low et al.,2020). Tools such as DALL-E, Gemini, Claude, and ChatGPT are transforming
business approaches to creativity and content development (Mackenzie,2024).
GenAI enables companies to automate content creation, personalize customer ex-
periences, and prototype ideas rapidly (García-Madurga & Grilló-Méndez,2023). Its
Adm. Sci. 2025,15, 494 https://doi.org/10.3390/admsci15120494
Adm. Sci. 2025,15, 494 2 of 23
transformative potential spans industries including marketing, design, tourism and en-
tertainment. Yet, its adoption is not without disruption, especially in SMEs, where stable
processes and limited resources dominate daily operations (Vial,2021). Integrating GenAI
challenges this stability, requiring strategic decisions about how to incorporate the tech-
nology: either through regenerative actions that reconfigure existing practices, or through
adaptive actions that integrate GenAI into current structures. These decisions are shaped by
organizational leadership (Haile,2023) behavior, which influences how managers perceive
change, mobilize resources, and guide their organizations through uncertainty in strategic
processes (Yukl,2012).
Despite growing interest, GenAI adoption among SMEs remains limited. In Spain, the
adoption of AI continues to increase, although it remains modest in comparative terms.
According to the latest report by the Observatorio Nacional de Tecnología y Sociedad
(ONTSI), 11.4% of companies with 10 or more employees used AI in 2024, representing
an increase of 1.8 percentage points compared with 2023 and more than 3 points relative
to 2021 (ONTSI,2025). Despite this upward trend, overall penetration remains limited,
particularly among SMEs, which continue to face resource constraints, lack of knowledge,
and organizational inertia that hinder broader adoption. These findings reveal the particular
difficulties SMEs face when adopting disruptive technologies. While extensive literature
exists on technology adoption, a critical gap remains in understanding how leadership
behavior influences the adoption of disruptive technologies like GenAI in SMEs digital
transformation efforts (Vial,2021). This is significant because organizational leadership
shapes the strategic alignment and digital transformation efforts must face the dynamic
and complex (Vial,2021) interactions among the relevant dimensions of the adoption
process (Venkatesh & Davis,2000;Barney,1991;Steel & König,2006) where multi-causal
relationships must be taken into consideration (Peñarroya-Farell et al.,2023). In this
vein, Organizational leadership must shift from a hierarchical to a more transformational
leadership approach that can accommodate the adaptability required to integrate such
technologies effectively. This study addresses this gap by exploring the role of Complexity
Leadership Theory (Uhl-Bien et al.,2007) in understanding organizational leadership
profiles to guide SMEs in strategic alignment for GenAI adoption.
It aims to generate insights into the interplay between leadership, strategic alignment,
and innovation management in the SME context.
With this purpose in mind, the research addresses the following questions:
1.
To what extent can an organizational leadership perspective shed light on disruptive
technology adoption in SMEs?
2.
To what extent can different organizational leadership profiles explain how SMEs
maintain competitiveness while aligning with strategic objectives during the adoption
of GenAI?
The next section presents the theoretical framework that underpins the research. By
integrating perspectives from technology adoption models, motivational theories, and
resource-based frameworks, the study aims to build a comprehensive understanding of
how organizational leadership profiles influence GenAI adoption in SMEs.
2. Theoretical Framework: Understanding GenAI Adoption in SMEs
The integration of Generative AI (GenAI) in SMEs represents a transformative shift
in business operations, yet its adoption remains complex (Babashahi et al.,2024). Existing
research has highlighted resistance to change, technical constraints, and strategic alignment
issues as key challenges (Dwivedi et al.,2024;Babashahi et al.,2024;Ooi et al.,2023). While
previous studies offer valuable insights into AI implementation, a clearer understanding is
needed on how SMEs navigate decisions around disruptive technologies like GenAI.
Adm. Sci. 2025,15, 494 3 of 23
2.1. Models of Technology Adoption in SMEs
Technology adoption in SMEs has been widely analyzed using models like the Technol-
ogy Acceptance Model (TAM) (Davis,1989), which emphasizes perceived usefulness and
ease of use (Venkatesh & Davis,2000). Extensions of TAM (e.g., TAM3) have introduced
contextual elements such as subjective norms and perceived risk (Venkatesh & Bala,2008).
However, TAM has been criticized for overlooking external factors such as competition
or regulation (Sun & Zhang,2006) and for its limited attention to organizational readiness
and leadership (Chau & Hu,2002).
Rogers (2003) proposed the Diffusion of Innovations theory, highlighting early
adopters and broad patterns of innovation diffusion. However, it pays less attention
to the internal decision-making processes and leadership dynamics within firms.
Temporal Motivation Theory (TMT) (Steel & König,2006) complements adoption models
by emphasizing urgency and motivational factors. SMEs may adopt GenAI when immediate
value is perceived (Khan & Khan,2024), but risks and uncertainties often delay action (Higgins,
2006;Steel,2007). Strategic planning mitigates these delays: well-defined transformation
strategies facilitate faster adoption (Grant & Ashford,2008), especially when motivational
triggers such as training and leadership support are present (Meyer et al.,2022).
From a resource perspective, the Resource-Based View (RBV) (Barney,1991) offers
insights into how internal capabilities like human capital and digital infrastructure support
adoption (Pfister & Lehmann,2022). Dynamic capabilities such as resource reconfiguration
(Teece et al.,1997;Eisenhardt & Martin,2000) are especially relevant when adopting
emerging technologies.
Consequently, GenAI can be studied based on the effects of each one of the previously
mentioned models; however, in dynamic, changing environments arising from the knowl-
edge era and digital transformation efforts, the interactions among technology adoption
frameworks, strategic resource management, and the temporal motivations embedded in
the models cannot be ignored (Venkatesh & Davis,2000;Barney,1991;Steel & König,2006).
For example, the motivation for technology implementation can influence perceptions
of ease of use or even the perceived usefulness of technology adoption (Venkatesh &
Bala,2008), and conversely, perceptions of adoption can influence motivation (Chau &
Hu,2002;Khan & Khan,2024). This mutual causality is also observed in the interaction
between technology adoption and the strategic vision of resources (Pfister & Lehmann,
2022) and, finally, between the strategic vision of resources and motivation (Rogers,2003;
Grant & Ashford,2008). In short, the usual linear models do not appear to be adequate for
environments such as that of digital transformation and disruptive technologies (Peñarroya-
Farell et al.,2023).
In this sense, this work assumes that integrating the three perspectives to achieve
strategic alignment in digital transformation actions, such as the adoption of GenAI by
SMEs, can be achieved through organizational leadership perspective (Haile,2023).
2.2. Organizational Leadership and GenAI Adoption in SMEs
The performance of GenAI adoption in SMEs must be linked to support organizational
goals. In this vein, from a dynamic capabilities’ lens (Teece,2018), alignment ensures co-
herence between intention, motivation, and resource allocation. Organizational leadership
(Haile,2023) has established that developing a business strategy requires an organizational
capacity that aligns the organization with strategic objectives. Organizational leadership is
a dynamic organizational capability derived from transformational leadership (Millar et al.,
2018), drawing on new leadership paradigms, such as distributed or shared leadership, and
linked to the Complexity Leadership Theory (CLT) (Uhl-Bien et al.,2007). This dynamic
Adm. Sci. 2025,15, 494 4 of 23
organizational capability (Haile,2023) defines leadership within the organization as a
management capacity that enables leaders to establish strategic objectives.
Organizational leadership must manage the complexity and uncertainty arising from the
multi-causal interactions among the three adoption perspectives, the inherent characteristics
of digital transformation, and the leadership actions involved in the process. This leadership
must be able to adapt dynamically to this interaction and establish the appropriate balance of
strategic priorities that should emerge from the adoption process. In this sense, the Complexity
Leadership Theory allows us to consider organizational leadership as a balance between
administrative, adaptive, and enabling processes (Uhl-Bien et al.,2007). It proposes a vision
of a Complex Adaptive System (CAS) that considers emerging situations, recombination
capabilities, and self-adaptation efforts, enabling a shift from hierarchical leadership to a
complex interplay between leaders and employees (Uhl-Bien et al.,2007).
2.3. A Complexity-Based Framework for SME Adoption
To address these limitations, this study integrates TAM, TMT, and RBV through the
lens of Complexity Theory (Holland,1992;Boulton et al.,2015) to propose a CAS model for
Organizational leadership.
Figure 1illustrates this complex system, placing strategic alignment as a key enabler
that connects individual perceptions (TAM), motivational urgency (TMT), and internal
capabilities (RBV). This holistic framework supports the interpretive analysis of SME
leadership behaviors, offering deeper insight into the nonlinear nature of GenAI adoption.
Figure 1. Complex system dynamics of Generative AI adoption in SMEs.
This model assumes that leadership in GenAI adoption can result in either an adaptive
or an enabling process (Boulton et al.,2015;Chiva et al.,2010;Holland,1992). In the first
case, the leader adapts the adoption to the company’s characteristics and does not prioritize
it. In the second case, the leader facilitates GenAI adoption and transforms the organization
for effective alignment with GenAI’s capabilities. In the adaptive approach, the TAM, TMT,
and RBV systems will behave in an adaptive manner, while in the enabling approach, all
three systems will behave in an enabling manner.
Adm. Sci. 2025,15, 494 5 of 23
This study presupposes that, contrary to linear adoption models, organizational
leadership profiles can emerge from a complex interplay among TAM, TMT, and RBV
systems that can lead to adoption behaviors that illustrate specific leadership profiles.
The next section outlines the study’s methodology, data collection, and analysis framework.
3. Methodology
This study examines how leadership behaviors influence GenAI adoption in SMEs
while maintaining strategic alignment. Using Interpretive Phenomenological Analysis (IPA)
(Eatough & Smith,2007), it explores how individuals interpret their experiences (Smith
et al.,2009) through a symbolic interactionist lens (Stryker,2008), identifying both shared
patterns and variations. To achieve this, a qualitative approach was employed to capture
the complexity of leadership behaviors and decision-making in GenAI adoption. Semi-
structured interviews provided rich, contextual insights, revealing patterns in leadership,
challenges, and strategic alignment. A cross-sectional design (Bughin & van Zeebroeck,
2017) offers a snapshot of SME leaders’ experiences amid rapid technological change. By
examining SMEs across various industries, the interpretivist approach explores leadership
perceptions of balancing innovation and stability.
3.1. Data Collection
Data was collected through semi-structured interviews from May to December 2024,
conducted face-to-face or via video conferencing based on participant availability. Each
session lasted 30–45 min, following a flexible interview guide. The guide, piloted with two
managers, included 20 questions exploring leadership strategies, challenges, and perspec-
tives on GenAI adoption while maintaining competitiveness and strategic alignment.
3.2. Sample and Participant Selection
The study used purposive and convenience sampling to recruit 15 strategic leaders
from Spanish SMEs across different industries. Participants were selected from an open
innovation workshop focused on the use of Generative AI in export and internationali-
sation functions. This recruitment context explains the predominance of CEOs, Export
Managers, and Commercial Directors in the final sample. These roles were deliberately
targeted because, in many SMEs, strategic technology- and innovation-related decisions
are concentrated in general management and commercial leadership rather than in spe-
cialised IT, operations, or finance departments. Consequently, although the sample is not
sector-balanced, it reflects the organisational structures typical of SMEs and captures the
viewpoints of those who most directly influence early GenAI adoption decisions in the
current development state of the adoption of GenAI. Early adoption stage.
Given the exploratory purpose of the study and the emergent state of GenAI adoption
in SMEs, a small-N, qualitative design was deemed appropriate. The sample size of 15 par-
ticipants aligns with qualitative research standards aimed at generating conceptual insight
rather than statistical generalisation. The goal was not to achieve industry-level satura-
tion but to identify cross-cutting leadership patterns and behavioural configurations that
emerge when SMEs begin engaging with GenAI. While the sample includes organisations
with different levels of digital maturity, its composition reflects the participant pool from
the workshop rather than a sector-stratified selection.
The over-representation of leaders from commercial and export functions introduces
an interpretive bias towards content creation, strategic communication, and customer-
facing applications of GenAI. However, this bias mirrors the reality that early adoption
within SMEs often originates in management or commercial areas, where experimentation
with new technologies is more feasible. This perspective was acknowledged as potentially
Adm. Sci. 2025,15, 494 6 of 23
underrepresenting GenAI applications in technical, operational, or financial processes, yet
it nevertheless provided valuable insight into how strategic leaders made sense of GenAI’s
potential and constraints within resource-limited environments.
Because the sample spans multiple industries yet remains small, its diversity should
be interpreted as providing conceptual breadth rather than sector-specific granularity. This
design supports the exploratory aim of identifying leadership-related mechanisms that
influence GenAI adoption under heterogeneous organisational conditions. As noted in
the limitations, the sectorally diffuse sample and the focus on roles linked to strategic and
internationalisation functions constrain the transferability of the findings. Future studies
could enhance the interpretive depth by incorporating operational and technical leaders
and by using sector-focused sampling strategies. (See Table 1).
Table 1. Overview of Interviewed Managers.
Subject Name Industry Position Key Characteristics
Manager 1 Healthcare service
provider
Innovation
Manager
Focuses on using AI for detailed data analysis and
strategic content creation, with a strong emphasis on
innovation to stay competitive.
Manager 2 AgriFood industry Export Manager Uses AI for process improvements, cautious
approach towards broader AI implementation.
Manager 3 Healthcare
manufacturer CEO AI is used for data analysis, content creation, and
internal efficiency improvements.
Manager 4 Food industry Export Manager Focuses on AI for translations, marketing, and
creating a Bot for export requirements.
Manager 5 Cosmetics
manufacturer CEO
Uses AI for demand forecasting and inventory
management, and cautious approach towards
broader AI implementation.
Manager 6 Industrial
Manufacturer
Owner &
Commercial Lead
AI is seen as a tool for quality control and process
optimization, facing organizational challenges
in adoption.
Manager 7 Programming Marketing
Manager
Top management shows some hesitation about
integrating GenAI (e.g., reluctance to share core
programming due to fears of confidentiality)
Manager 8 Online marketplace CEO
AI is used to create communication content in
diverse formats (text, video, images), but it is not
incorporated in other departments.
Manager 9 International
Trading company CEO
The manager uses generative AI to ask questions as
if it were Google, but it is more complex. It has not
incorporated it into the company’s processes.
Manager 10 Translation Services CEO
The company is adapting to market changes by
exploring GenAI to improve efficiency and offer
innovative services, while maintaining its core focus
on technical translation and localization.
Manager 11 Food and
beverages Wholesaler Export Manager
While they are actively exploring the potential of
GenAI, their adoption is driven by immediate
practical needs with a cautious approach to
broader transformation.
Manager 12 Laboratory
manufacturer CEO They have recently begun exploring the use of AI,
applying it on a limited basis to create content.
Adm. Sci. 2025,15, 494 7 of 23
Table 1. Cont.
Subject Name Industry Position Key Characteristics
Manager 13 Electronics
Wholesaler CEO They are using GenAI for diverse
administrative tasks.
Manager 14 Online e-Commerce CEO
Uses GenAI to create product images and is
beginning to integrate it into the company’s
marketing strategy.
Manager 15 B2B Services Marketing
Manager
Uses GenAI at a personal level but not within the
company, as there is a lack of trust in its reliability.
In addition to the variables reported in Table 1, contextual information was collected
to support interpretation of the findings. This included the approximate size of each
firm (number of employees) and an assessment of its digital maturity level, derived from
participants’ descriptions of existing tools, processes, and prior digital transformation
initiatives. These contextual descriptors were not central to the study’s aims and are not
presented in table format, but they were considered during the interpretive process to avoid
overgeneralisation and to ensure that leadership behaviours were contextualised within
each firm’s structural and technological conditions.
3.3. Data Analysis Tool
Data were analysed using a qualitative, interpretive approach informed by elements
of Interpretative Phenomenological Analysis (IPA) (Smith et al.,2009), combined with
thematic analysis techniques (Braun & Clarke,2006). While IPA traditionally prioritises
in-depth, idiographic examination of a small number of homogeneous participants, the
present study adopted a partial and adapted use of IPA. Specifically, IPA was employed
to guide the interpretive orientation of the analysis, focusing on how SME leaders make
sense of GenAI adoption, while thematic analysis supported the identification of cross-case
patterns across a heterogeneous sample.
Each interview was first analysed individually to preserve the idiographic sensitivity
characteristic of IPA, allowing the researchers to capture the personal meaning-making
processes expressed by each participant. Subsequently, codes and emergent concepts were
compared across cases to identify shared themes. This configurational, cross-case stage
departs from a strict IPA design but was considered appropriate given the exploratory aims
of the study and the need to understand leadership behaviour across SMEs operating in
diverse contexts.
This hybrid analytical strategy aligns with the study’s objective of uncovering leader-
ship patterns rather than conducting a purely phenomenological reconstruction of each
individual case. The approach preserves the interpretive depth of IPA while enabling a
broader thematic synthesis. As acknowledged in the limitations section, this methodolog-
ical adaptation represents a partial fit with IPA’s idiographic principles and reduces the
depth traditionally associated with full IPA, but it provides an analytically coherent foun-
dation for exploring leadership behaviours in early-stage GenAI adoption among SMEs.
Given the interpretive nature of the study, reflexivity was incorporated throughout the
analytical process. The lead researcher and one of the coauthors have extensive professional
experience in digital transformation and AI adoption in SMEs, which offers informed
sensitivity but also potential biases in interpreting participants’ accounts. To mitigate
this risk, reflexive notes were kept during coding, interpretations were cross-checked
among co-authors, and participants were invited to review interview summaries to validate
Adm. Sci. 2025,15, 494 8 of 23
accuracy. These procedures aimed to ensure transparency and reduce the influence of prior
assumptions on the development of themes and leadership patterns.
3.4. Data Analysis
Using thematic analysis (Braun & Clarke,2006), recurring topics were classified within
the study’s theoretical framework. The key categories identified include Perceived Useful-
ness, Perceived Ease of Use, Attitude Toward Using, Behavioral Intention to Use, Exter-
nal Variables, Perceived Value of GenAI, Urgency of Adoption, Capability Building and
Training, and Infrastructure Readiness. Figure 2illustrates the nine categories and their re-
lationship with the three main themes: Intention, Motivation, and Resource Allocation. As
organizations function as complex systems (Anderson,1999), several categories influence
more than one theme, reflecting the interconnected nature of AI adoption. Solid lines indi-
cate primary relationships, while dashed lines represent indirect influences, underscoring
the dynamic and interdependent nature of decision-making. In line with recent literature,
these interdependencies also reflect processes of mutual causality among technology adop-
tion constructs, motivational mechanisms, and resource-based conditions (Venkatesh &
Bala,2008;Chau & Hu,2002;Pfister & Lehmann,2022). Motivation can shape perceptions
of usefulness and ease of use, just as these perceptions can enhance or inhibit motivation;
similarly, resource strategies influence motivational readiness and are themselves shaped
by evolving adoption practices. Such reciprocal interactions highlight that linear models
are insufficient for explaining adoption dynamics in contexts characterised by digital trans-
formation and disruptive technologies, reinforcing the need for a complexity-informed
interpretation of the data.
Figure 2. Thematic map of key themes and their relationships. Solid lines indicate primary relation-
ships, while dashed lines represent indirect influences.
Adm. Sci. 2025,15, 494 9 of 23
4. Findings
4.1. Intention to Use GenAI Strategically in Their Companies
The intention analysis identified four key categories: (a) Attitude toward using GenAI,
(b) Perceived ease of use, (c) Perceived usefulness, and (d) Behavioral intention to use. The
results reveal a divide between leaders who view GenAI as a strategic necessity and those
who remain doubtful or hesitant about its integration. A significant portion of leaders
(f13, where f represents frequency) consider GenAI fundamental for business efficiency,
reflecting strong confidence in its potential to enhance productivity and competitiveness.
“We see GenAI as a fundamental tool for managing large amounts of data efficiently.”
(Manager 8)
“It will be as transformative as YouTube was for learning. GenAI adoption will grow
exponentially as more people realize its value.” (Manager 4)
Manager 4 also views GenAI as “the next generational shift akin to the Industrial Revolu-
tion,” reflecting her strong belief in its transformative potential. However, a considerable
number of managers (f9) express skepticism, suggesting uncertainty about its practical
benefits or feasibility.
“The company has a strong innovation culture, implementing major technological ad-
vances. However, we are hesitant to adopt GenAI due to fears of losing control over
proprietary information.” (Manager 7)
Some leaders (f4) take a more cautious stance, viewing GenAI merely as a complement
rather than a transformative tool, which may limit its adoption to supporting rather than
reshaping business processes. Others confess using it at a personal level but not strategically
for their companies (f3).
“At a personal level, I have used ChatGPT, but we have not implemented it in the
company.” (Manager 2)
Additionally, resistance from senior employees (f3) indicates internal barriers that
could hinder implementation, emphasizing the impact of organizational culture and work-
force adaptation on the adoption process. Age appears to be a factor in this resistance, as
highlighted by Manager 12:
“Younger employees, particularly in marketing and sales, are more open to it.”
(Manager 12)
Younger professionals, especially in tech-driven roles, are more likely to embrace
GenAI, while older employees may resist due to unfamiliarity, usability concerns
(Venkatesh et al.,2003), risk aversion (Morris & Venkatesh,2000;Babashahi et al.,2024),
and job security fears (Bélanger & Carter,2008). Rogers’ Diffusion of innovations theory
(Rogers,2003) suggests younger employees adopt new technologies faster, whereas older
workers benefit from structured training and leadership support (Zacher & Rosing,2015).
Adoption success depends on balancing accessibility with proper education and oversight,
as some find GenAI intuitive (f7), while others see a need for supervision (f5) or additional
training (f5).
“For our daily use, AI tools are intuitive. Once we understand how to phrase queries, the
experience becomes seamless.” (Manager 3)
“Yes, it’s intuitive. The only challenge is verifying accuracy, as GenAI sometimes
generates misleading information.” (Manager 11)
The need for learning may also reflect the evolving nature of GenAI tools, requiring
continuous adaptation as capabilities expand. Additionally, access to advanced training
Adm. Sci. 2025,15, 494 10 of 23
materials provided by GenAI developers themselves may present a linguistic barrier.
Manager 13 highlighted this challenge, noting that most documentation and learning
resources for Gemini AI were only available in English:
“Much of the technical information is only available in English, making the learning
curve steeper.” (Manager 13)
This suggests that non-English-speaking users may face additional obstacles when
trying to deepen their expertise, potentially slowing down adoption and limiting the full
exploitation of GenAI’s capabilities.
Interestingly, it was observed that those who used GenAI less frequently found it
easier to use, while more experienced users felt they needed further training. This aligns
with the Dunning-Kruger Effect (Kruger & Dunning,1999), where lower-competence
individuals overestimate their abilities, while experts recognize knowledge gaps. Rogers’
(2003) diffusion of innovations theory similarly suggests that early adopters may not fully
grasp a technology’s complexities until deeper engagement reveals challenges.
The most cited GenAI application was content creation (f12), widely used for market-
ing, emails, and reports. Translations (f6) and marketing and sales (f6) also emerged as key
uses, particularly in multilingual and international business contexts.
“I use GenAI extensively in export operations. It helps with translations, email structur-
ing, and generating marketing ideas.” (Manager 4)
Beyond these primary functions, GenAI is also recognized for its role in automating
administrative tasks (f3), supporting market research (f3), and managing documents (f2),
reflecting its utility in enhancing efficiency by handling repetitive processes and structuring
business information. Some businesses have also begun exploring its potential in customer
support (f1), financial analysis (f1), and e-Commerce (f1), while others see value in its
ability to aid in business presentations (f1), idea validation (f1), and decision-making
(f1). However, the data also reveals some hesitation regarding its reliability. While many
acknowledge its usefulness, one participant explicitly noted that GenAI is “useful but not
reliable” (f1) (Manager 9), suggesting concerns about accuracy or consistency, while another
stated that they saw “no use in production” (f1) (Manager 5), indicating skepticism about its
applicability in core business operations.
“GenAI is useful for administrative tasks, but when it comes to the production floor, I
simply don’t trust it and I don’t see how it can help us.” (Manager 5)
The findings indicate that while GenAI enhances content creation and efficiency,
trust in its output’s limits adoption in strategic areas. Many businesses (f6) report no
formal AI implementation, reflecting uncertainty, resource constraints, or a lack of strategic
direction. In some cases, employees drive adoption (f5), experimenting with GenAI without
direct leadership involvement. This bottom-up adoption suggests a leadership gap, where
managers either overlook GenAI’s potential or hesitate to integrate it into strategy. Manager
2, an export manager in agrifood, exemplifies this—using GenAI personally while higher
management resists innovation, relying on traditional, paper-based methods.
For some companies, financial constraints hinder adoption, with some seeking funding
for implementation (f2) and others for training (f2). While these businesses see GenAI’s
potential, cost remains a major barrier to scaling efforts. Findings suggest that interest
in GenAI is high, but structured adoption is limited. Employee-driven use highlights a
grassroots approach, while financial constraints slow broader adoption. For some SMEs,
investment in infrastructure and training is essential to move beyond experimental use.
Adm. Sci. 2025,15, 494 11 of 23
4.2. Motivation to Use GenAI in Their Companies
The motivation analysis revealed three major categories: (a) External variables af-
fecting the implementation, (b) Perceived Value of GenAI, and (c) Urgency to implement
GenAI. Despite the growing discourse around GenAI, no participant reported feeling
pressured by competitors to implement it (f15). While media coverage had prompted some
to experiment with the technology (f4), this was driven by curiosity rather than competitive
necessity. Most respondents doubted that their competitors were using GenAI beyond
basic content creation, and many believed that even this application was not widespread.
“I’ve noticed better-written texts from my competitors on LinkedIn, but I doubt they’re
using GenAI for anything else.” (Manager 1)
The lack of perceived competitive pressure suggests that businesses do not yet see
GenAI as a critical differentiator, potentially underestimating its long-term impact. How-
ever, Manager 10, CEO of the translation services, showcased the decline in business due
to GenAI-based translation solutions.
“We have seen a decline in business as some clients have shifted to GenAI-based translation
solutions instead of using our services.” (Manager 10)
This reflects a competitive shift driven by client behavior rather than industry peers,
indicating that some sectors may experience disruption sooner than others. This aligns with
Christensen’s (1997,2013) Disruptive Innovation Theory, which explains how emerging
technologies can displace traditional service providers by offering cheaper and more scal-
able alternatives. In this case, GenAI acts as a substitute technology (Bower & Christensen,
1995), replacing human translation for many routine tasks, thereby reducing the demand
for professional translators. As a result, businesses in the translation sector must reconsider
their strategic positioning.
While many SMEs do not yet perceive direct competition from GenAI adopters,
market dynamics and changing customer expectations could accelerate the urgency for
adoption, particularly in industries where GenAI serves as a direct substitute rather than
a complement. The most cited benefit is improving efficiency (f8), showing that decision-
makers view GenAI as a tool to streamline operations and reduce workload. Boosting
productivity (f5) follows, reinforcing AI’s role in increasing output without added costs.
“It saves me significant time, especially in structuring emails and generating multilingual
campaigns.” (Manager 4)
Improving creativity (f2) is mentioned less often, suggesting that SMEs focus more on
GenAI’s practical applications than on its potential for innovation. Similarly, competitive
advantage (f1) and internal transformation (f1) are rarely cited, indicating that most leaders
see GenAI as a tool for incremental improvements rather than business model innovation.
These findings suggest that the primary motivation for adopting GenAI in SMEs is
rooted in efficiency and productivity gains, rather than in transformative change. While
a few companies see GenAI as a competitive differentiator, most seem to perceive it as
a supporting tool rather than a strategic game-changer. This aligns with broader trends
in SME digital adoption, where resource constraints and risk management often lead to
gradual and pragmatic implementation rather than radical shifts in business operations
(Mohamed & Weber,2020).
Personal motivation (f7) is the primary driver of GenAI adoption, with external
pressures playing a lesser role. Many participants do not see GenAI as an immediate
priority (f 5), though some acknowledge growing urgency (f4), aligning with Temporal
Motivation Theory (Steel & König,2006), which links urgency to perceived proximity and
expected value. Few leaders (f2), such as Manager 10 in translation services, see GenAI as a
Adm. Sci. 2025,15, 494 12 of 23
necessity, aligning with Perceived Strategic Urgency Theory (Eisenhardt & Martin,2000),
where adoption is critical for survival.
Findings suggest GenAI adoption in SMEs remains voluntary, not necessity driven.
Prospect Theory (Kahneman & Tversky,1979) explains this reluctance: firms act more on
imminent threats than potential gains. While urgency is rising for some, most SMEs still
view GenAI as optional, potentially slowing widespread adoption.
4.3. Resource Allocation for GenAI in SMEs
The analysis of the Resource Distribution for GenAI revealed two major categories:
(a) Resource Mobilization for Capability Building and Training, and (b) Infrastructure
Readiness. The most frequently mentioned need is general GenAI training (f10), suggesting
that many companies recognize a knowledge gap and seek foundational understanding.
Additionally, specialized training in Sales and Marketing (f5) and International Expansion
(f5) indicates a demand for GenAI education tailored to specific business functions.
“I recently reviewed an GenAI training course, but it seemed too technical for our sales
team, so I didn’t proceed with it.” (Manager 12)
Meanwhile, a few companies are seeking funding for training (f4), reinforcing resource
constraints may limit structured learning opportunities. GenAI automation (f1) appears
as a niche training need, indicating that only a few businesses are considering deeper
automation strategies. While training needs are widely acknowledged, many SMEs rely
on self-learning through YouTube videos and other informal resources, as systematic and
structured training programs remain scarce.
“I am starting to take some free online courses to better understand how it works.”
(Manager 15)
This suggests that while interest in GenAI education is growing, access to high-quality,
tailored training remains a challenge for SMEs, potentially slowing adoption and effective
implementation. Infrastructure Readiness (IR) refers to the extent to which an organization
possesses the necessary technological resources to effectively integrate and utilize artificial
intelligence. This includes elements such as computing power, data storage, networking
capabilities, and specialized GenAI tools. Within this framework, paid GenAI is not merely
an auxiliary function but a critical component of a company’s infrastructure. The data
reveals that a majority of businesses (f7) are not yet ready, indicating that technological
limitations, lack of IT support, or outdated systems may hinder adoption. However, some
companies (f5) are actively working on readiness, suggesting a transition phase where
businesses recognize the need for upgrades but have not yet fully implemented them. Only
a small number of SMEs (f3) report being fully prepared, highlighting that few businesses
have the necessary infrastructure in place for seamless GenAI integration.
These findings suggest infrastructure is a major bottleneck in GenAI adoption, with
technical barriers slowing implementation despite high motivation and perceived useful-
ness. Limited IT resources and digital transformation challenges hinder SMEs’ ability to
leverage GenAI effectively. From a Resource-Based View (Barney,1991), digital infrastruc-
ture is a strategic asset, but financial constraints and a lack of AI talent (Rietmann,2021)
create a digital divide, favoring well-resourced firms. Without targeted investment, SMEs
risk falling behind in AI-driven markets.
5. Discussion
5.1. Leadership Behavior and Disruptive Technology Adoption in SMEs
Addressing Research Question 1, the findings indicate that GenAI adoption in SMEs
arises from the interaction among leaders’ intention, motivation, and resource alloca-
Adm. Sci. 2025,15, 494 13 of 23
tion, rather than from their linear influence. This supports complexity-based views of
organizational change, which conceptualize digital transformation as a nonlinear and
interdependent process (Anderson,1999;Uhl-Bien et al.,2007). Leadership behavior shapes
GenAI adoption not by acting on isolated drivers but by influencing how these dimensions
evolve together as part of a complex adaptive system.
From a technology acceptance perspective, perceived usefulness and ease of use (Davis,
1989;Venkatesh & Davis,2000) do play a role, but the findings show that positive percep-
tions alone are insufficient. Their effect depends on the firm’s resource base and dynamic
capabilities (Barney,1991;Teece et al.,1997;Vial,2021). Leaders often recognize GenAI’s
value yet postpone adoption due to limited skills, IT readiness, or budget constraints,
illustrating how TAM variables become conditioned by structural factors emphasized in
the RBV.
Motivational mechanisms also contribute to this configuration. In line with Temporal
Motivation Theory (Steel & König,2006;Steel,2007), urgency and perceived value interact with
leaders’ risk perceptions. When urgency is low, even strong positive attitudes do not lead to
action; when urgency increases, due to market pressures or operational needs—adoption may
occur despite capability gaps. These findings reflect the temporal trade-offs that characterize
SME decision-making (Higgins,2006;Bughin & van Zeebroeck,2017).
Complexity Leadership Theory offers an integrative explanation. Leaders operate
across administrative, adaptive, and enabling roles (Uhl-Bien et al.,2007), and GenAI
adoption is more advanced in firms where leadership fosters experimentation, learning,
and capability development. Conversely, when leadership favors stability and control,
adoption remains incremental or symbolic. This highlights that leadership influences not
only attitudes or resources but the alignment among intention, motivation, and capabilities,
which is essential in dynamic digital environments (Vial,2021;Teece,2018).
Overall, the results show that leadership behavior in SMEs functions as a coordinating
mechanism that shapes the interplay of cognitive (TAM), motivational (TMT), and resource-
based (RBV) elements. GenAI adoption becomes viable when these elements reinforce
each other, rather than operate independently. This extends traditional adoption models
by positioning leadership as the central factor that determines whether GenAI remains a
peripheral tool or evolves into a driver of strategic renewal.
5.2. Leadership Profiles and the Dynamics of GenAI Adoption in SMEs
Table 2provides the first integrative overview of how each manager in the study
approaches GenAI across the nine analytical categories identified earlier in the thematic
analysis. These categories, covering intention elements (Perceived Usefulness, Perceived
Ease of Use, Attitude, Behavioral Intention), motivation elements (External Variables,
Perceived Value, Urgency), and resource elements (Capability-Building and Infrastructure
Readiness), are coded as either adaptive or regenerative. This classification reflects whether
leaders engage with GenAI through incremental, risk-sensitive behaviors or through a
future-oriented, transformative stance grounded in exploration and capability expansion.
It brings together the three theoretical lenses guiding the study. From a TAM per-
spective (Davis,1989;Venkatesh & Davis,2000), adaptive vs. regenerative variation in
usefulness and ease-of-use perceptions helps explain differences in technological intention
and acceptance. The TMT dimension (Steel & König,2006) is visible in how managers’ ur-
gency and value perceptions combine to either motivate or delay adoption. Finally, the RBV
perspective (Barney,1991) is evident in the resource categories, where capability-building
and infrastructure readiness impose clear constraints that shape leaders’ overall orientation.
On the other hand, it shows that no manager demonstrates a fully regenerative profile
across all nine categories. Even leaders with strong transformative intentions often show
Adm. Sci. 2025,15, 494 14 of 23
adaptive behavior in resource-related categories, highlighting capability gaps common in
SMEs (Pfister & Lehmann,2022;Vial,2021). This suggests that enthusiasm for GenAI does
not automatically translate into investment, readiness, or strategic commitment, a finding
consistent with early-stage diffusion conditions.
Table 2. Leadership approach classification.
Intention Motivation Resource Al.
PU PEU ATT BEH EV PV URG CBT IR
M1 REGEN ADPT REGEN REGEN ADPT REGEN ADPT REGEN REGEN
M2 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M3 REGEN ADPT REGEN REGEN ADPT ADPT ADPT REGEN ADPT
M4 REGEN ADPT REGEN REGEN ADPT REGEN ADPT REGEN REGEN
M5 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M6 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M7 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M8 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M9 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M10 REGEN ADPT REGEN REGEN ADPT ADPT REGEN REGEN REGEN
M11 REGEN ADPT REGEN REGEN ADPT ADPT ADPT ADPT ADPT
M12 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
M13 REGEN ADPT REGEN REGEN ADPT REGEN ADPT ADPT REGEN
M14 REGEN ADPT REGEN REGEN ADPT ADPT ADPT ADPT ADPT
M15 ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT ADPT
Ultimately, Table 2serves as the analytical bridge between the individual-level coding
and the four broader leadership patterns synthesized later in Table 6. By displaying
how adaptive and regenerative behaviors cluster differently among managers, it becomes
possible to trace how distinct configurations of intention, motivation, and resources give rise
to the Strategic, Aspiring, Opportunistic, and Operational leadership profiles. It therefore
plays a foundational role in transitioning from thematic codes to configuration-based
theoretical interpretation.
The emergence of the four leadership profiles, Strategic Adopters, Aspiring Adopters,
Opportunistic Adopters, and Operational Stabilizers, reflects the configurational interplay
among intention (Table 3), motivation (Table 4), and resource allocation (Table 5). Rather
than aligning with the sequential adopter categories proposed by Rogers (2003), these pro-
files arise from the non-linear, multi-causal interactions between leaders’ perceptions rooted
in TAM (Davis,1989;Venkatesh & Davis,2000), their assessments of urgency and value
consistent with TMT (Steel & König,2006;Steel,2007), and the organization’s capabilities as
described by the RBV and dynamic capabilities perspectives (Barney,1991;Teece et al.,1997;
Teece,2018). Drawing on Complexity Theory (Anderson,1999) and Complexity Leadership
Theory (Uhl-Bien et al.,2007), these leadership patterns can be understood as emergent
states within a complex adaptive system, where leadership behavior, resource constraints,
and perceived opportunities co-evolve. This lens helps explain not only why these four
profiles form, but also how they stabilize or shift over time as feedback loops between
intention, motivation, and capability development reinforce or alter adoption trajectories.
Strategic Adopters exhibit regenerative perceptions and behaviors across all three
tables: high perceived usefulness and ease of use (Table 3), strong value-based motivation
and urgency (Table 4), and proactive capability-building and infrastructure investment
(Table 5). While these leaders resemble “innovators” in Rogers’ diffusion model, their
trajectories are better explained by dynamic capabilities theory (Teece et al.,1997;Teece,
2018). Early investment reduces uncertainty, strengthens perceived value (Davis,1989;
Adm. Sci. 2025,15, 494 15 of 23
Venkatesh & Davis,2000), and creates reinforcing loops in which resources, learning, and
motivation amplify one another. Complexity theory conceptualizes this configuration
as a high-change attractor, where experimentation and capability development reinforce
strategic renewal.
Table 3. Summary of Intention Category (TAM-Aligned Variables).
Intention (TAM) PU PEU ATT BEH Manager
MID REGEN_IN REGEN ADPT REGEN REGEN
M1, M3, M4, M10, M11, M13, M14
FULL ADPT_IN ADPT ADPT ADPT ADPT M2, M5–M9, M12, M15
Note: Includes Perceived Usefulness (PU), Perceived Ease of Use (PEU), Attitude (ATT), and Behavioral Intention
(BEH), summarizing leader intention configurations.
Table 4. Summary of Motivation Category (TMT-Aligned Variables).
Motivation (TMT) EV PV URG Manager
MID ADPT1_MOT ADPT REGEN ADPT M1, M4, M13
MID ADPT2_MOT ADPT ADPT REGEN M10
FULL ADPT_MOT ADPT ADPT ADPT M2–M3, M5–M9, M11–M12; M14–M15
Note: Summarizes External Variables (EV), Perceived Value (PV), and Urgency (URG) to classify motivational
configurations.
Table 5. Summary of Resource Allocation Category (RBV-Aligned Variables).
Resource Al. (RBV) CBT IR Manager
FULL REGEN_RA REGEN REGEN M1, M4, M10
MID ADPT1_RA REGEN ADPT M3
MID ADPT2_RA ADPT REGEN M13
FULL ADPT_RA ADPT ADPT M2, M5-M9, M11–M12; M14–M15
Note: Includes Capability-Building (CBT) and Infrastructure Readiness (IR), which represent resource-related
configurations.
Aspiring Adopters show regenerative intention in Table 3but stronger constraints in
motivational drivers (Table 4) and resource allocation decisions (Table 5). This creates a
resource–motivation feedback loop (Steel & König,2006;Higgins,2006): limited resources
reduce urgency, weakened urgency delays investment, and capability gaps reinforce hes-
itation (Barney,1991;Pfister & Lehmann,2022). Unlike Rogers’ “early majority,” who
progress through predictable stages, these leaders operate within a fragile, path-dependent
equilibrium. Complexity theory suggests that small interventions, policy incentives, mar-
ket pressures, and access to shared digital infrastructures can shift their pattern toward
Strategic Adopters, while persistent constraints can lock them into stagnation.
Opportunistic Adopters present adaptive intention (Table 3) and tactical motivation
(Table 4), with selective and localized investment decisions (Table 5). Their behavior is
less aligned with Rogers’ “early majority” and more consistent with local optimization
within complex systems: immediate operational needs, symbolic pressures, or short-term
opportunities shape adoption choices (Bughin & van Zeebroeck,2017;Vial,2021). Because
these choices are not anchored in strategic alignment, they yield fragmented innovation,
with isolated experiments seldom scaling into organization-wide transformation. Com-
plexity theory describes this as a shallow adoption attractor, where movement is dynamic
but non-transformative unless leaders link operational experimentation to strategic learn-
ing processes.
Operational Stabilizers exhibit adaptive intention, low motivational intensity, and mini-
mal resource commitment across the three tables. While superficially resembling Rogers’ “late
Adm. Sci. 2025,15, 494 16 of 23
adopters,” their behavior is grounded in path dependence and self-reinforcing constraints:
low perceived usefulness suppresses motivation, limited investment undermines IT readiness,
and weak readiness validates initial perceptions (García-Madurga & Grilló-Méndez,2023).
Complexity theory frames this as a low-change attractor state, where the system stabilizes
around non-adoption. Only significant external shocks, regulatory changes, competitive
pressures, or structural policy interventions are likely to break this cycle.
Table 6synthesizes how intention (Table 3), motivation (Table 4), and resources
(Table 5) combine into the four leadership profiles. Importantly, these profiles should
not be interpreted as fixed categories or natural stages. Complexity theory highlights that
leadership configurations behave as dynamic states: firms may shift from one profile to
another as perceptions evolve, resources are reconfigured, or environmental pressures
intensify. This configurational view extends classical models, TAM (Davis,1989), TMT
(Steel & König,2006), RBV (Barney,1991), and Rogers’ diffusion theory, by showing that
SME GenAI adoption emerges through iterative feedback loops rather than linear, pre-
dictable sequences.
Table 6. Integrated Leadership Profiles (Configurational Patterns Across Tables 3–5).
Pattern Intention Motivation Resource Al Manager
Strategic Adopters MID REGEN_IN MID ADPT_MOT FULL REGEN_RA M1, M4, M10
Aspiring Adopters MID REGEN_IN FULL ADPT_MOT MID ADPT_RA M3, M13
Opportunistic Adopters
MID REGEN_IN FULL ADPT_MOT FULL ADPT_RA M11, M14
Operational Stabilizers FULL ADPT_IN FULL ADPT_MOT FULL ADPT_RA M2, M5-M9, M12, M15
Note: Synthesizes intention, motivation, and resource configurations into four emergent leadership profiles:
Strategic, Aspiring, Opportunistic, and Operational.
The configurational analysis presented in Tables 3–6and Figures 2–5shows that the
four leadership profiles emerge from the interaction of intention, motivation, and resource
allocation. These patterns are not linear stages of adoption, nor do they align with Rogers’
(2003) classical adopter categories. Instead, they represent complex adaptive states (An-
derson,1999), shaped by ongoing feedback between leadership behavior, organizational
conditions, and emerging opportunities. Viewed through Complexity Leadership Theory
(Uhl-Bien et al.,2007), these profiles illustrate how leadership actions and resource con-
straints co-evolve over time, producing distinct and dynamic pathways for GenAI adoption
in SMEs.
Pattern 1: Strategic Adopters. Figure 3visualizes a configuration where regenerative
intention, strong motivation, and proactive resource allocation co-evolve. Although these
leaders resemble Rogers’ “innovators,” their behavior is best understood through the lens of
dynamic capabilities (Teece et al.,1997;Teece,2018). Early investments reduce uncertainty,
strengthen perceived usefulness and ease of use (Davis,1989;Venkatesh & Davis,2000),
and create positive reinforcement cycles. The values correspond to the leadership approach
in each dimension: 4 = Full Regenerative, 3 = Mid Regenerative, 2 = Mid Adaptive, and
1 = Full Adaptive. Complexity theory conceptualizes this as a high-change attractor,
where experimentation, capability-building, and strategic alignment reinforce one another,
accelerating transformation.
Pattern 2: Aspiring Adopters. Figure 4illustrates a leadership configuration defined
by regenerative intention but constrained motivation and capability gaps. This pattern
reflects a resource–motivation feedback loop (Steel & König,2006;Higgins,2006): limited
resources suppress urgency, lower urgency delays investment, and delayed investment
reinforces the capability deficit (Barney,1991;Pfister & Lehmann,2022). Complexity theory
explains this as a fragile equilibrium, a system easily shifted by small triggers (external
incentives, client demands, shared tools) but equally prone to stagnation if constraints
Adm. Sci. 2025,15, 494 17 of 23
remain unresolved. This pattern does not directly correspond to Rogers’ framework but
emerges as a distinctive leadership behavior within the complex systems perspective. Their
presence highlights that GenAI adoption in SMEs is not purely a matter of intention but
also of systemic readiness and resource mobilization (Figure 4).
Figure 3. Strategic innovator s pattern.
Figure 4. Aspiring adopter’s pattern.
Pattern 3: Opportunistic Adopters. These leaders exhibit a high intention to adopt
GenAI but lack intrinsic motivation and do not allocate sufficient resources for its imple-
mentation. Figure 5depicts leaders who apply GenAI tactically to localized operational
tasks rather than as part of strategic transformation. Rather than aligning with Rogers’
“early majority,” their behavior reflects local-optimization dynamics typical of complex
systems. They pursue small, low-risk wins shaped by immediate pressures (Bughin & van
Zeebroeck,2017;Vial,2021), but these isolated experiments rarely scale unless intentionally
linked to organizational learning. This forms a shallow adoption attractor, where innova-
tion is dynamic but lacks depth and integration. Rather than being true pioneers of digital
transformation, these organizations engage in what could be considered symbolic adop-
tion, demonstrating interest in GenAI without the necessary groundwork for meaningful
integration. As a result, their efforts may stagnate, reinforcing the paradox where digital
investments fail to translate into tangible competitive advantages (Figure 5).
Pattern 4: Operational Stabilizers. These leaders follow a fully adaptive approach.
They do not prioritize GenAI as a driver of change; instead, they focus on maintaining oper-
ational efficiency with minimal disruption. Figure 6represents a configuration dominated
by adaptive intention, low motivation, and minimal resource allocation. Although similar
to late adopters, their inertia is driven less by resistance and more by path dependence
and reinforcing loops: low perceived usefulness reduces motivation to invest, limited
investment depresses IT readiness, and low readiness confirms leaders’ initial perceptions
(García-Madurga & Grilló-Méndez,2023). Complexity theory frames this as a low-change
Adm. Sci. 2025,15, 494 18 of 23
attractor, where small interventions have limited effect unless accompanied by external
shocks or significant structural support.
Figure 5. Opportunistic adopter’s pattern.
Figure 6. Operational Stabilizers” pattern.
While Rogers’ diffusion model remains a useful foundation, it lacks the nuance to
fully capture the diversity of leadership behavior in GenAI adoption. Our analysis shows
that early-stage adopters are not homogeneous, and crossing Moore (1991) Chasm requires
more than technological readiness; it requires strong leadership, strategic clarity, and
capability development.
Leadership is not a passive enabler, but an active force shaping the trajectory and
depth of GenAI integration. Leaders who align intention, motivation, and resources are
more likely to move beyond superficial use and unlock GenAI’s full potential. Conversely,
leaders who delay investment or underestimate strategic alignment risk falling into the
“digitalization paradox”, where adoption fails to deliver value.
Understanding these differentiated leadership behaviors equips practitioners and
policymakers to better support SMEs by tailoring interventions to their specific stage and
readiness in the adoption journey.
Beyond explaining how these configurations emerge, the typology also carries impor-
tant managerial and theory-testing implications. From an organizational perspective, each
profile reflects a distinct readiness path: Strategic Adopters illustrate how early investment
and clear value articulation accelerate adoption; Aspiring Adopters demonstrate the need
for targeted capability-building to overcome motivational and resource bottlenecks; Op-
portunistic Adopters highlight the risks of fragmented experimentation without strategic
alignment; and Operational Stabilizers show how path dependence can limit technological
renewal unless externally prompted. These profiles, therefore, help managers diagnose
their organization’s current position and identify levers, such as urgency framing, skill de-
Adm. Sci. 2025,15, 494 19 of 23
velopment, or infrastructural upgrading, that could shift them toward more transformative
trajectories. From a theoretical standpoint, the typology provides a foundation for future
testing by offering empirically grounded configurations that can be operationalized using
TAM, TMT, and RBV constructs and validated through comparative designs such as fsQCA
or larger-scale quantitative studies. This underscores the typology’s potential not only as
an interpretive lens but as a framework for cumulative theory building in the context of
GenAI adoption in SMEs.
6. Conclusions
This study examined how SME leaders approach the adoption of Generative AI within
the broader digital transformation challenges outlined in the Introduction. Consistent
with prior work emphasizing the disruptive nature of GenAI and the structural limitations
that SMEs face (Dwivedi et al.,2024;ONTSI,2025;Vial,2021), the findings show that
adoption emerges from the interdependence of leaders’ intentions, motivational dynamics,
and resource structures. By integrating the Technology Acceptance Model (Davis,1989;
Venkatesh & Davis,2000), Temporal Motivation Theory (Steel & König,2006), the Resource-
Based View (Barney,1991), dynamic capabilities (Teece et al.,1997;Teece,2018), and
complexity theory (Anderson,1999), the study provides a holistic interpretation of how
these elements interact in real SME environments.
The analysis identified four leadership profiles: Strategic Adopters, Aspiring Adopters,
Opportunistic Adopters, and Operational Stabilizers, which illustrate that GenAI adoption
does not follow linear or sequential pathways but emerges as a set of dynamic config-
urations shaped by feedback loops between leadership behaviour, perceived opportu-
nities, and capability constraints. This perspective aligns with Complexity Leadership
Theory (Uhl-Bien et al.,2007) and contributes theoretically by extending classical adop-
tion models through an understanding of non-linear, emergent adoption dynamics in
resource-constrained contexts.
From a managerial standpoint, the implications of the study are intentionally high-
level and should be interpreted as orientational rather than prescriptive, given the ex-
ploratory qualitative design. The findings indicate that progression in GenAI adoption
depends on leaders’ ability to articulate a clear strategic purpose, cultivate a realistic sense
of urgency, and foster basic internal capabilities while working within constrained techno-
logical and financial resources (Vial,2021;Teece,2018). Rather than pursuing advanced or
disruptive applications from the outset, SMEs appear to benefit from small-scale experi-
mentation, structured learning, and incremental capability-building that reduce uncertainty
and encourage employee engagement. Leadership behaviours that support knowledge
sharing, enable experimentation, and integrate GenAI initiatives with existing resources
can help shift SMEs from fragmented or ad hoc use to more coherent, strategically aligned
adoption patterns, even when resource limitations remain significant. These insights re-
flect the adaptive, emergent practices observed among the SMEs in this study and align
with the complexity-based understanding of digital transformation developed throughout
the paper.
The study’s conclusions also reflect the influence of sample composition. Because par-
ticipants were primarily managers in commercial, marketing, and internationalisation roles,
the dominant use cases identified were content creation, translation, and communication
tasks. This pattern should not be interpreted as evidence that GenAI’s value is limited to
these functions. As highlighted in the Introduction and supported by existing research,
GenAI also holds significant potential for forecasting, optimisation, and production-related
decision-making (Hofmann et al.,2019;Ivanov & Dolgui,2021), but these operational
contexts were not represented in the dataset.
Adm. Sci. 2025,15, 494 20 of 23
Overall, this study positions GenAI adoption in SMEs as a context-dependent, non-
linear, and evolving process shaped by leadership configurations that emerge through
the interaction of intention, motivation, and resource allocation. While the implications
must be viewed within the constraints of an exploratory qualitative design, the findings
offer a foundation for future research aimed at tracing how leadership configurations shift
over time, how sectoral context conditions adoption dynamics, and how SMEs develop
the dynamic capabilities needed to integrate GenAI more strategically and sustainably
(Figure 7).
Figure 7. Diverse pathways to GenAI adoption in SMEs.
Study Limitations and Suggestions for Future Research
This study is exploratory and relies on a qualitative, small-N design, which limits
the transferability and generalizability of the findings (Smith et al.,2009;Pietkiewicz &
Smith,2014). As already clarified in Section 3, the study employed a partial and adapted
use of Interpretative Phenomenological Analysis (IPA). This adaptation departs from IPA’s
idiographic and homogeneous-sample requirements and prioritizes cross-case pattern
identification over full phenomenological depth. Such a methodological choice represents a
formal limitation, as it affects the interpretive coherence and idiographic richness typically
expected in IPA studies, and should be taken into account when interpreting the emergent
leadership profiles.
A second limitation concerns the heterogeneity and diffuse nature of the small sample,
which included leaders from diverse sectors, functions, and levels of digital maturity. While
this heterogeneity enabled the identification of broad leadership configurations, it may also
have introduced interpretive bias. In particular, the predominance of participants from
commercial, marketing, and internationalisation roles likely contributed to the prominence
of content-creation use cases observed in the findings. This sectoral bias constrains the
study’s ability to speak to GenAI adoption in operational, manufacturing, logistics, or
supply-chain contexts, where GenAI may support forecasting, optimisation, or anomaly
detection (Hofmann et al.,2019;Ivanov & Dolgui,2021). Future studies should consider
more targeted, sector-specific sampling strategies to enhance interpretive depth and reduce
functional bias.
Adm. Sci. 2025,15, 494 21 of 23
A third limitation relates to the study’s geographical focus. The sample consists solely
of SME leaders operating within the Spanish context, which may limit the applicability of
the findings to different institutional, cultural, or regulatory environments. Cross-national
comparative work would allow for a deeper understanding of how leadership configurations
interact with varying digital maturity levels and policy frameworks (Vial,2021).
These limitations open several avenues for future research directly connected to the
study’s methodological constraints. Future studies could adopt longitudinal qualitative
designs to examine how leadership configurations evolve over time as SMEs progress
through different stages of GenAI adoption. Alternatively, comparative or mixed-methods
approaches, including fsQCA or causal-comparative designs, could provide stronger insight
into the configurational conditions under which specific leadership profiles emerge. Finally,
sector-focused samples and idiographically coherent IPA studies would allow researchers to
deepen understanding of leadership sense-making in specific organizational environments
while preserving IPA’s methodological integrity.
Author Contributions: Conceptualization, M.P.-F., M.V. and F.M.; Methodology, M.P.-F., M.V.
and S.K.S.R.; Software, M.P.-F. and M.V.; Validation, M.P.-F. and S.K.S.R.; Formal analysis, M.P.-
F. and S.K.S.R.; Investigation, M.P.-F. and M.V.; Resources, M.P.-F.; Data curation, M.P.-F. and M.V.;
Writing—original draft, M.P.-F. and F.M.; Writing—review & editing, M.P.-F., M.V., S.K.S.R. and F.M.;
Visualization, M.P.-F. and M.V.; Supervision, F.M.; Project administration, F.M.; Funding acquisition,
F.M. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: According to the ethical procedures of La Salle—Universitat
Ramon Llull (Barcelona, Spain), this type of non-experimental qualitative research—based on
anonymized voluntary interviews and not involving the collection of personal or sensitive data—does
not require formal approval from the university’s Ethics Committee.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: The qualitative data used in this study consist of transcribed interviews
that contain contextual and potentially identifiable information. For privacy and ethical reasons,
these data cannot be made publicly available. However, the authors are willing to provide access to
anonymized excerpts or coding outputs upon reasonable request.
Conflicts of Interest: The authors declare no conflict of interest.
References
Anderson, P. (1999). Complexity theory and organization science. Organization Science,10(3), 216–232. [CrossRef]
Babashahi, L., Barbosa, C. E., Lima, Y., Lyra, A., Salazar, H., Argôlo, M., de Almeida, M. A., & de Souza, J. M. (2024). AI in the
workplace: A systematic review of skill transformation in the industry. Administrative Sciences,14(6), 127. [CrossRef]
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management,17(1), 99–120. [CrossRef]
Bélanger, F., & Carter, L. (2008). Trust and risk in e-government adoption. Journal of Strategic Information Systems,17(2), 165–176.
[CrossRef]
Boulton, J. G., Allen, P. M., & Bowman, C. (2015). Embracing complexity: Strategic perspectives for an age of turbulence. OUP Oxford.
Bower, J. L., & Christensen, C. M. (1995). Disruptive technologies: Catching the wave. Harvard Business Review,73(1), 43–53.
Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology,3(2), 77–101. [CrossRef]
Bughin, J., & van Zeebroeck, N. (2017). The best response to digital disruption. MIT Sloan.
Chau, P. Y., & Hu, P. J. (2002). Investigating healthcare professionals’ decisions to accept telemedicine technology: An empirical test of
competing theories. Information & Management,39(4), 297–311. [CrossRef]
Chiva, R., Grandío, A., & Alegre, J. (2010). Adaptive and generative learning: Implications from complexity theories. International
Journal of Management Reviews,12(2), 114–129. [CrossRef]
Christensen, C. M. (1997). The innovator’s dilemma: When new technologies cause great firms to fail. Harvard Business School Press.
Christensen, C. M. (2013). The innovator’s solution: Creating and sustaining successful growth. Harvard Business Review Press.
Adm. Sci. 2025,15, 494 22 of 23
Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly,13(3),
319–340. [CrossRef]
Dwivedi, Y. K., Pandey, N., Currie, W., & Micu, A. (2024). Leveraging ChatGPT and other generative artificial intelligence (AI)-
based applications in the hospitality and tourism industry: Practices, challenges, and research agenda. International Journal of
Contemporary Hospitality Management,36(1), 1–12. [CrossRef]
Eatough, V., & Smith, J. A. (2007). Interpretative phenomenological analysis. In C. Willig, & W. Stainton-Rogers (Eds.), The SAGE
handbook of qualitative research in psychology (pp. 179–194). Sage.
Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: What are they? Strategic Management Journal,21(10–11), 1105–1121.
[CrossRef]
García-Madurga, M.-Á., & Grilló-Méndez, A.-J. (2023). Artificial intelligence in the tourism industry: An overview of reviews.
Administrative Sciences,13, 172. [CrossRef]
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., & Bengio, Y. (2020). Generative adversarial
networks. Communications of the ACM,63(11), 139–144. [CrossRef]
Grant, A. M., & Ashford, S. J. (2008). The dynamics of proactivity at work. Research in Organizational Behavior,28, 3–34. [CrossRef]
Haile, G. A. (2023). Organizational leadership: How much does it matter? British Journal of Industrial Relations,61(3), 653–673. [CrossRef]
Higgins, E. T. (2006). Value from hedonic experience and engagement. Psychological Review,113(3), 439–460. [CrossRef] [PubMed]
Hofmann, E., Sternberg, H., Chen, H., Pflaum, A., & Prockl, G. (2019). Supply chain management and Industry 4.0: Conducting
research in the digital age. International Journal of Physical Distribution & Logistics Management,49(10), 945–955. [CrossRef]
Holland, J. H. (1992). Complex adaptive systems. Daedalus,121(1), 17–30.
Ivanov, D., & Dolgui, A. (2021). A digital supply chain twin for managing the disruption risks and resilience in the era of Industry 4.0.
Production Planning & Control,32(9), 775–788. [CrossRef]
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica,47(2), 263–292. [CrossRef]
Khan, U., & Khan, K. A. (2024). Generative artificial intelligence in hospitality and tourism marketing: Perceptions, risks, benefits, and
policy implications. Journal of Global Hospitality and Tourism,3(1), 269–284.
Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it: How difficulties in recognizing one’s own incompetence lead to inflated
self-assessments. Journal of Personality and Social Psychology,77(6), 1121–1134. [CrossRef] [PubMed]
Mackenzie, A. H. (2024). Artificial intelligence: Friend or foe? The Science Teacher,91(2), 6–7. [CrossRef]
Meyer, J., de Reuver, M., & Bouwman, H. (2022). Motivational triggers in SME digital transformation: A mixed-methods study. Journal
of Small Business & Enterprise Development,29(3), 403–426.
Millar, C. C., Groth, O., & Mahon, J. F. (2018). Management innovation in a VUCA world: Challenges and recommendations. California
Management Review,61(1), 5–14. [CrossRef]
Mohamed, N., & Weber, R. (2020). Digital transformation in SMEs: The role of resources and financial constraints in adoption. arXiv,
arXiv:2002.11623.
Moore, G. A. (1991). Crossing the chasm: Marketing and selling high-tech products to mainstream customers. Harper Business.
Morris, M. G., & Venkatesh, V. (2000). Age differences in technology adoption decisions: Implications for a changing workforce.
Personnel Psychology,53(2), 375–403. [CrossRef]
ONTSI. (2025). Indicadores del uso de inteligencia artificial en España 2024 (Edición 2025–Datos 2024). Observatorio Nacional de Tecnología
y Sociedad. Available online: https://www.ontsi.es (accessed on 7 December 2025).
Ooi, K.-B., Tan, G. W.-H., Al-Emran, M., Al-Sharafi, M. A., Capatina, A., Chakraborty, A., Dwivedi, Y. K., Huang, T.-L., Kar, A. K., Lee,
V.-H., Loh, X.-M., Micu, A., Mikalef, P., Mogaji, E., Pandey, N., Raman, R., Rana, N. P., Sarker, P., Sharma, A.,
. . .
Wong, L.-W.
(2023). The potential of generative artificial intelligence across disciplines: Perspectives and future directions. Journal of Computer
Information Systems,65, 76–107. [CrossRef]
Peñarroya-Farell, M., Miralles, F., & Vaziri, M. (2023). Open and sustainable business model innovation: An intention-based perspective
from the Spanish cultural firms. Journal of Open Innovation: Technology, Market, and Complexity,9(2), 100036. [CrossRef]
Pfister, P., & Lehmann, C. (2022). Digital value creation in German SMEs—A return-on-investment analysis. Journal of Small Business &
Entrepreneurship,36(4), 548–573. [CrossRef]
Pietkiewicz, I., & Smith, J. A. (2014). A practical guide to using interpretative phenomenological analysis in qualitative research
psychology. Czasopismo Psychologiczne—Psychological Journal,20(1), 7–14. [CrossRef]
Rietmann, C. (2021). Digital pioneers in the periphery? Toward a typology of rural hidden champions in times of digitalization. Journal
of Small Business & Entrepreneurship,36(2), 281–301. [CrossRef]
Rogers, E. M. (2003). Diffusion of innovations. Free Press.
Sagala, G. H., & ˝
Ori, D. (2024). Exploring digital transformation strategy to achieve SMEs resilience and antifragility: A systematic
literature review. Journal of Small Business and Entrepreneurship,37, 495–524. [CrossRef]
Smith, J. A., Flowers, P., & Larkin, M. (2009). Interpretative phenomenological analysis: Theory, method and research. Sage.
Adm. Sci. 2025,15, 494 23 of 23
Steel, P. (2007). The nature of procrastination: A meta-analytic and theoretical review of quintessential self-regulatory failure.
Psychological Bulletin,133(1), 65. [CrossRef]
Steel, P., & König, C. J. (2006). Integrating theories of motivation. Academy of Management Review,31(4), 889–913. [CrossRef]
Stryker, S. (2008). From Mead to a structural symbolic interactionism and beyond. Annual Review of Sociology,34(1), 15–31. [CrossRef]
Sun, H., & Zhang, P. (2006). The role of moderating factors in user technology acceptance. International Journal of Human-Computer
Studies,64(2), 53–78. [CrossRef]
Teece, D. J. (2018). Dynamic capabilities as (workable) management systems theory. Journal of Management & Organization,24(3),
359–368. [CrossRef]
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal,18(7),
509–533. [CrossRef]
Uhl-Bien, M., Marion, R., & McKelvey, B. (2007). Complexity leadership theory: Shifting leadership from the industrial age to the
knowledge era. The Leadership Quarterly,18(4), 298–318. [CrossRef]
Venkatesh, V., & Bala, H. (2008). Technology acceptance model 3 and a research agenda on interventions. Decision Sciences,39(2),
273–315. [CrossRef]
Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies.
Management Science,46(2), 186–204. [CrossRef]
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view.
MIS Quarterly,27, 425–478. [CrossRef]
Vial, G. (2021). Understanding digital transformation: A review and a research agenda. In Managing digital transformation (pp. 13–66).
Routledge.
Yukl, G. (2012). Effective leadership behavior: What we know and what questions need more attention. Academy of Management
Perspectives,26(4), 66–85. [CrossRef]
Zacher, H., & Rosing, K. (2015). Ambidextrous leadership and team innovation. Leadership & Organization Development Journal,36(1),
54–68. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
