Starting Point

International
The value of AI on entrepreneurship:
Journal of
Entrepreneurial
evidence from the European Union
Behavior &
Research
Hasanul Banna
Department of Finance and Economics, Manchester Metropolitan University,
Manchester, UK and
Miyan Research Institute, International University of Business Agriculture and
Received 31 December 2024
Technology, Dhaka, Bangladesh, and
Revised 9 June 2025
15 October 2025
Ashraful Alam
24 November 2025
Accepted 25 November 2025
Salford Business School, University of Salford, Manchester, UK
Abstract
Purpose – This study explores the impact of artificial intelligence (AI) investments on entrepreneurship,
focusing on their influence on firm revenue growth across 26 European countries from 2012 to 2023. The
research aims to uncover the dynamics of AI investment, particularly its short-term challenges and long-term
benefits and to examine the critical interplay between AI adoption and R&D innovation strategies.
Design/methodology/approach – The analysis uses an unbalanced panel dataset of 1,479 firms, applying the
Cameron et al. (2011) multi-way clustering (CGM) estimation technique to account for heteroscedasticity and
cross-sectional dependence. Robustness tests include alternative AI proxies and instrumental variable (2SLS-
IV) regression to mitigate potential endogeneity issues.
Findings – The results reveal a U-shaped relationship between AI investments and revenue growth, indicating
that initial AI adoption may hinder revenue growth due to high upfront costs or inefficiencies. However,
significant long-term revenue benefits emerge as firms gain experience with AI. Additionally, integrating AI
with innovation strategies substantially enhances revenue growth, highlighting that standalone AI investments
are insufficient for achieving entrepreneurial success.
Originality/value – This study contributes to the literature by providing empirical evidence on the dual-phase
impact of AI investments on firm performance and emphasising the strategic importance of aligning AI adoption
with innovation efforts. The findings offer actionable insights for policymakers and business leaders aiming to
leverage AI for sustained entrepreneurial growth.
Keywords Artificial intelligence (AI), Entrepreneurship, Innovation, R&D, Resource-based theory
Paper type Research article
1. Introduction
Artificial intelligence (AI) is attracting significant attention from entrepreneurs due to its data-
driven insights and predictive capabilities. According to Statista (2025), the AI market is
estimated to have reached $244 billion in 2025, reflecting a substantial increase of nearly $58
billion compared to 2024. Entrepreneurs with innovative mindsets are leveraging AI to identify
and create new business opportunities, informed decision-making, performance, and education
and research (Giuggioli and Pellegrini, 2023). While AI systems can help address challenges
stemming from uncertainty through future predictions (Townsend et al., 2025a, b), they also
exhibit fundamental limitations. Nambisan et al. (2019) argued that AI struggles to process
entirely new datasets and generate truly original, out-of-the-box ideas. Since entrepreneurs often
experiment with novel concepts, there is limited evidence regarding AI’s role in entrepreneurial
activities and the revenue it generates. Moreover, the ease of access and relatively affordable cost
have made AI solutions practical and no longer mere futuristic innovations (Iansiti and Lakhani,
2020; Giuggioli and Pellegrini, 2023). As a result, the value of AI investments for entrepreneurs
remains uncertain, prompting us to study this area more closely.
Businesses must assess the value of AI investments to determine whether integrating AI
makes financial sense. By understanding AI’s potential, entrepreneurs can strategically
International Journal of Entrepreneurial
Behavior & Research
allocate resources to ensure that AI investments maximise their impact on customer
© Emerald Publishing Limited
experience, productivity, and long-term growth. Clark (2024) highlighted that AI tools can e-ISSN: 1758-6534
p-ISSN: 1355-2554
create value by reducing costs and enhancing profitability. However, not all firms fully benefit DOI 10.1108/IJEBR-12-2024-1489
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

from AI due to insufficient capabilities or integration. For instance, small firms often face
IJEBR
challenges such as the high costs of AI software, hardware, infrastructure, and the additional
burden of learning how to use AI tools effectively, compounded by financial and staffing
constraints. As a result, the relationship between AI and entrepreneurial innovation may
plateau or even show negative effects. This suggests that the conditions under which AI
investments deliver real benefits are still unclear. In this context, Jia et al. (2024) argued that
combining AI with a firm’s innovation capacity can help unlock its potential value.
Accordingly, this study investigates to what extent and under what conditions AI investments
contribute to entrepreneurial firm growth.
By doing so, this study makes several contributions to the literature. First, it adds to the
growing body of research examining the relationship between AI and entrepreneurship.
However, there is limited research on how and when AI investments can generate value for
entrepreneurial activities. This study complements the work of Giuggioli and Pellegrini (2023)
and Chen and Zhang (2023), who explored the role of AI in entrepreneurship. While Chen and
Zhang (2023) focused on China, this study extends the research to European countries.
In contrast to Giuggioli and Pellegrini’s (2023) systematic literature review, our study provides
empirical evidence on the relationship between AI and entrepreneurship. This study finds that
there is a U-shaped relationship between AI investment and entrepreneurial activities.
Recognising a U-shaped relationship helps entrepreneurs identify the tipping point, the level of
AI investment where benefits start to outweigh the costs. It ensures they avoid under or over
investing in AI technologies.
Second, this paper contributes to the extensive literature on entrepreneurial outcomes.
It has been argued that merely investing in AI is not sufficient to achieve competitive
advantages. To fully realise the benefits of AI, firms need to integrate AI adoption with
innovation capabilities. Previous research, such as Mariani et al. (2023) and Calabro et al.
(2019), which employed bibliometric and literature review analyses, suggested that AI
adoption must align with innovation to maximise its potential. This research extends their
studies by examining the interaction between AI investments and R&D activities to evaluate
whether this integration yields fruitful outcomes. This study differs by providing quantitative
evidence that captures AI adoption over time and its outcomes in a longitudinal context. The
study reveals that combining AI investments with active R&D efforts better harnesses AI’s
innovative potential, leading to improved entrepreneurial outcomes. Furthermore, the results
show that firm size plays an important role in the interaction between AI adoption and R&D
activities of the firm. In particular, small firms potentially benefit more from AI investments
over time than larger firms. These findings emphasise the need for a nuanced understanding of
how firm size influences the interaction between AI adoption and R&D. Small firms can
leverage AI as a strategic tool for long-term growth, while larger firms may need to innovate
their integration strategies to avoid diminishing returns. This insight has broad implications for
managers, policymakers, and researchers aiming to optimise the benefits of AI across firms of
varying sizes.
The remainder of the paper is structured as follows: Section 2 outlines the theoretical
framework and hypothesis development, Section 3 details the methodology, data, and sample,
Section 4 presents the findings, and Section 5 provides the conclusion.
2. Theory and hypotheses development
2.1 Literature review
AI encompasses a broad spectrum of computing techniques capable of performing tasks
traditionally requiring human intelligence, such as natural language processing, computer
vision, and machine learning applications including chatbots, text generation, facial
recognition, autonomous driving, and recommendation systems (Lui et al., 2022). Unlike
conventional software that follows predetermined rules, AI systems can learn from data, adapt
to new information, and improve performance over time (Chalmers et al., 2021). This adaptive
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

capacity has encouraged an increasing number of organisations to adopt AI for prediction, International
analysis, risk management, and strategic decision-making. For instance, chatbots and Journal of
voicebots now manage substantial portions of customer service functions across retail, Entrepreneurial
Behavior &
banking, and healthcare, while financial institutions employ AI to detect fraud and predict loan
Research
defaults (Naveed, 2022; Aziz and Andriansyah, 2023). Evidence shows that firms investing in
AI often achieve higher growth in sales, employment, and market valuation, as AI facilitates
innovation and the creation of new products and services (Babina et al., 2024). Moreover, AI
enhances organisational productivity by reducing errors, automating routine tasks, and
enabling employees to focus on higher value activities (Lui et al., 2022).
For entrepreneurs, AI offers both major opportunities and complex challenges. It can
help them find new markets, make data-driven decisions, and improve innovation and
efficiency (Giuggioli and Pellegrini, 2023; Iansiti and Lakhani, 2020; Townsend et al.,
2025a, b). AI can also lower barriers to entry, simplify administrative tasks, and help
businesses grow more easily (OECD, 2025). However, achieving these benefits is not easy.
Many firms face high costs of setting up and maintaining AI systems (Kraus et al., 2021),
lack skilled staff to manage them (Dicuonzo et al., 2023), and struggle with data-related
issues (Sun and Medaglia, 2019). There are also concerns about job losses and
overdependence on machines (Cubric, 2020), as well as the absence of clear strategies
for adopting AI effectively (Kar et al., 2021). Consequently, while AI holds substantial
potential to enhance firm performance, its adoption remains uneven due to financial,
technical, and organisational constraints.
Although empirical evidence supports the performance benefits of AI (Giuggioli and
Pellegrini, 2023; Babina et al., 2024; Pham et al., 2024), much of the existing literature
assumes a simple, linear relationship between AI adoption and performance, overlooking
possible non-linear or lagged effects. In reality, the performance impact of new technologies
often unfolds gradually as firms learn and adapt. Historical evidence from earlier technological
transitions supports this view: Brynjolfsson et al. (1994) found that the strongest
organisational impacts of IT investments appeared only after a two-to three-year lag, while
Loveman (1994) reported that IT investments had no immediate effect on output. These
findings highlight the well-known “productivity paradox”, in which technology adoption
initially generates adjustment costs and inefficiencies before delivering measurable gains.
Such insights suggest that, like IT, the benefits of AI depend heavily on a firm’s innovative
capacity, absorptive capability, and its ability to integrate AI effectively into organisational
routines and decision-making processes. Consequently, the relationship between AI adoption
and firm performance is dynamic rather than static, evolving over time as firms reconfigure
their resources and learn to exploit AI’s full potential. Furthermore, the generalisability of
existing research remains limited, as many prior studies are conceptual or confined to specific
national contexts. For instance, Chen and Zhang (2023) focus on China, while Pham et al.
(2024) and Babina et al. (2024) examine the United States; Chalmers et al. (2021), in contrast,
offer a conceptual discussion without empirical validation.
Previous studies show that AI can improve business performance, but the size and direction
of its impact can differ across countries and over time. Yet, there is little evidence on how this
relationship works in Europe, where business environments and innovation systems are
different from those in the United States or Asia. Most existing research has also ignored the
possibility that the effect of AI may not be straightforward or immediate but could follow a
non-linear pattern. In addition, few studies have explored how AI and research and
development (R&D) investment work together to support business growth. This study
addresses these gaps by examining the non-linear link between AI adoption and firm growth in
European countries and by analysing how R&D investment strengthens the benefits of AI for
innovation and performance. In doing so, it moves beyond earlier studies that focus on single
countries or assume simple linear effects, offering a clearer picture of how AI supports
business success in different European contexts.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

2.2 Theoretical framework and hypotheses
IJEBR
Resource Based Theory (RBT) suggests that a firm’s resources and capabilities are critical for
achieving competitive advantage and sustaining long-term success (Barney, 2001).
Organisational resources serve as the inputs that enable a company to perform its activities,
achieve goals, and create value (Madhani, 2010). According to RBT, resources must be
Valuable, Rare, Inimitable and Non-substitutable (VRIN) to generate and maintain sustainable
competitive advantage. This principle highlights that advantage arises not simply from owning
resources, but from developing and protecting those that are distinctive and difficult for
competitors to replicate. However, RBT has been criticised for its static nature and its limited
ability to explain how resources are developed, renewed, or reconfigured in response to
changing conditions that underpin competitive advantage (Easterby-Smith and Prieto, 2008;
Priem and Butler, 2001). In the context of digital transformation, such as the adoption of AI,
success depends not only on the possession of valuable resources but also on the capabilities
required to sustain competitive advantage. For example, firms may invest in AI technologies,
but without the necessary learning routines, absorptive capacity, or integration skills, these
resources may fail to deliver long-term value. This limitation demonstrates the need to
understand not only what resources a firm possesses but how they are used and adapted in
practice.
To address these limitations, Teece et al. (1997) introduced the Dynamic Capabilities
Theory (DCT), which extends RBT by emphasising a firm’s ability to respond with resilience
and agility to a changing environment. DCT identifies three key mechanisms: sensing, seizing,
and reconfiguring that enable firms to identify new opportunities, mobilise resources to
capture them, and transform their resource base to remain competitive. In order to effectively
integrate AI technologies and translate them into entrepreneurial innovation, dynamic
capabilities such as innovation capability, organisational learning, and agile decision-making
are essential. This integration between RBT and DCT provides a more comprehensive
theoretical foundation, where VRIN resources form the basis of advantage, while dynamic
capabilities ensure these resources are continually renewed and aligned with environmental
change.
AI adoption acts as an enabler of dynamic capabilities, allowing firms to become more
adaptive to environmental and technological change (Wong and Ngai, 2025). AI has the
potential to enhance the capabilities of a company’s human resources (Przegalinska et al.,
2025) by automating routine tasks, supporting decision-making, and enabling employees to
focus on higher value, creative, and strategic activities. In addition, the use of AI artefacts helps
firms to interpret complex data, uncover patterns, and make strategic decisions based on these
insights (Dubey et al., 2021). As a result, Davenport and Ronanki (2018) noted that with the
significant growth in data and computing power, AI is playing an increasingly important role
in improving business performance and driving competitive advantage. Viewed through the
lens of DCT, AI enhances a firm’s sensing capability by enabling data-driven opportunity
recognition, strengthens seizing by supporting timely strategic responses, and improves
reconfiguring by facilitating ongoing process renewal and innovation. Using AI, firms are
therefore able to identify new opportunities, develop innovative solutions, and increase
operational efficiency (Wong and Ngai, 2025).
While AI offers promising potential for enhancing firm performance, its adoption also
presents substantial challenges. Firms often face issues such as misalignment between AI
applications and managerial objectives, risks of operational errors or reputational damage, and
significant implementation and transition costs that require organisational restructuring and
employee upskilling (Lui et al., 2022; Bughin et al., 2018). As a result, the effects of AI
investment are not uniformly positive. AI adoption demands substantial financial resources
and carries considerable uncertainty regarding its benefits and payback period (Lui et al.,
2022). PwC (2015) found that higher spending on technology does not automatically translate
into improved profitability, suggesting that investment alone is insufficient to generate value.
Similarly, research on innovation and IT capital shows that returns may initially be weak or
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

negative until a critical threshold is reached, after which performance improves (Jen and International
Ju, 2005). Journal of
In the early stages, AI adoption may even hinder performance due to adjustment costs, skill Entrepreneurial
Behavior &
gaps, and learning inefficiencies (Dwivedi et al., 2021). Brynjolfsson (1993) noted that
Research
technological learning takes time, as firms must accumulate experience before fully realising
productivity gains. Initial investments often result in higher costs than benefits, but as
organisations progress along the learning curve, efficiency improves and long-term returns
begin to materialise. Consistent with this, Santos and Sussman (2000) observed that IT related
benefits often emerge only after a delay, while Brynjolfsson (1993) similarly found that IT
investments exhibit lagged effects on performance.
These insights suggest that the relationship between AI investment and firm growth is
unlikely to be linear. At low levels of investment, the costs and inefficiencies of adoption may
outweigh the benefits, leading to a temporary decline in performance. However, as firms
accumulate experience, optimise AI integration, and strengthen their innovation capabilities,
the marginal returns on AI investment can become increasingly positive. Ultimately, effective
AI deployment can enhance operational efficiency, speed, and cost reduction, thereby
stimulating revenue growth (Ghazwani et al., 2022). For instance, Rayner (1995) found that IT
investments accounted for as much as 9% of revenue in some industries, underscoring the
strategic importance of technological investment once maturity is achieved. Based on the
above discussion, this study proposes the following hypothesis:
H1. AI investment has a U-shaped effect on revenue growth, such that initial increases in
AI investment reduce growth, but beyond a certain point, further investment
enhances growth.
Investment in research and development (R&D) is a key indicator of a firm’s innovativeness.
Higher R&D intensity not only signals a strong commitment of resources to innovation but
also builds absorptive capacity, enabling the firm to better absorb and utilise new knowledge
and technology. Cohen and Levinthal (1990) stated that when firms acquire knowledge from
external sources, how well it complements their existing internal knowledge depends on their
absorptive capacity, their ability to recognise, assimilate, and apply new knowledge. Prior
studies (Suhyeon et al., 2023; Calabro et al., 2019; Kang and Kang, 2009) also demonstrate
that firm innovativeness or creativity largely depends on the ability to absorb external
knowledge.
Innovation can be further enhanced through the integration of artificial intelligence (AI),
which supports firms in generating and evaluating new ideas, conducting analyses, and
improving decision making processes. AI empowers humans to enhance their skills, complete
tasks more efficiently, and achieve superior outcomes, thereby fostering greater creativity and
innovation (Wilson and Daugherty, 2018). Moreover, AI has the potential to revolutionise
innovation management by facilitating more effective and efficient innovation processes
(Fu€ller et al., 2022). To strengthen their capabilities and maintain a competitive edge,
companies are increasingly incorporating AI technologies into their innovation strategies
(Bahoo et al., 2023).
Firms that combine AI investments with robust R&D efforts are better positioned to unlock
AI’s innovative potential, yielding enhanced entrepreneurial outcomes. Therefore, it is
assumed that firms with strong R&D capabilities can absorb AI technology more effectively.
In a recent study, Jia et al. (2024) pointed out that integrating AI with a company’s innovation
efforts can foster new processes and boost employee creativity. This synergy enables
organisations to develop novel products and processes that drive growth more effectively than
relying on either AI or R&D alone. By leveraging R&D based absorptive capacity, firms can
translate AI driven insights into commercially viable innovations, thereby converting
technological capability into measurable performance gains such as revenue growth. Based on
the above discussion, the study posits the following hypothesis:
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

H2. The interaction between AI investments and R&D positively influences changes in
IJEBR
revenue growth.
3. Method
3.1 Sample
The data for this research is sourced from multiple high-quality databases. Firm level financial
data is obtained from S&P Capital IQ, providing comprehensive insights into corporate
financials. Data on Artificial Intelligence (AI) is retrieved from the OECD AI Policy
Observatory database. Macroeconomic indicators are sourced from the World Bank’s World
Development Indicators (WDI), while institutional quality metrics are derived from the World
Governance Indicators (WGI).
Initially, the study considers a sample of 27 European Union (EU) member countries,
including Austria, Belgium, Bulgaria, Croatia, Cyprus, Czechia, Denmark, Estonia, Finland,
France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta,
Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden. However,
due to the unavailability of AI-related data, Lithuania is excluded, resulting in a final sample of
26 EU countries. The focus on the EU leverages its unique combination of regulatory
frameworks, coordinated innovation policies (e.g. Horizon Europe, Digital Europe), economic
diversity, and high-quality data availability, providing an ideal setting to study the relationship
between AI investments and entrepreneurship. The insights gained from this study not only
enhance the understanding of EU-specific dynamics but also provide a distinct and
underexplored lens compared to US centric studies, allowing for generalisable insights across
varying institutional and market conditions.
From these 26 EU countries, an initial dataset of 6,585 publicly listed firms is identified.
Firms without R&D expenditure data are excluded, reducing the sample size to 1,807 firms.
Further, firms with fewer than three consecutive years of financial data are removed, excluding
an additional 328 firms. After filtering missing data, the final unbalanced panel data comprises
1,479 firms and 12,942 firm-year observations. The analysis spans the period from 2012 to
2023, aligning with the availability of AI related data across all variables. We match firm year
observations with country level AI investment data using the firm’s primary country of
incorporation, following a refined variant of the methodology used in Felten et al. (2021). This
approach improves upon prior work by (1) covering a larger number of firms across multiple
countries and sectors, and (2) enhancing temporal granularity through year level matching.
The sample encompasses firms from a diverse array of industries, including communication
services, consumer discretionary, consumer staples, energy, financials, health care, industrials,
information technology, materials, real estate, and utilities. To ensure that extreme values do
not unduly influence the analysis, all data are winsorised at the 1st and 99th percentiles.
The firms analysed in this study are characterised as entrepreneurial due to their adherence
to core principles of entrepreneurship. They engage in sustained innovation by committing to
R&D investments and incorporating advanced technologies like AI, aligning with
Schumpeter’s (1934) view of entrepreneurial activity as market transformation driven by
innovation (Schumpeter and Swedberg, 2021). In line with Gali et al. (2024), only firms with
significant R&D expenditures (averaging USD 345 million annually within the sample) were
included to ensure the focus remained on innovation driven enterprises. Moreover, revenue
growth, utilised as the dependent variable, serves as a critical metric of entrepreneurial
success, reflecting the ability of these firms to leverage innovation for competitive advantage
(Davidsson and Wiklund, 2001). Additionally, their operations within the EU, a region known
for promoting entrepreneurship through frameworks like Horizon Europe, strengthen the
entrepreneurial context of this study. Furthermore, these firms span diverse, innovation
intensive sectors such as information technology and healthcare, showcasing adaptability and
strategic management of market and technological uncertainties. This aligns with Sarasvathy’s
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

(2001) effectuation theory, which emphasises entrepreneurial resilience and proactive International
decision-making in uncertain environments. Journal of
Firms are classified as entrepreneurial when they exhibit at least two of three empirically Entrepreneurial
Behavior &
measurable and externally verifiable characteristics, such as innovation intensity, growth
Research
dynamism, and entrepreneurial financing, within a rolling five-year observation window. Each
dimension is benchmarked relative to industry–year–size normalised z-scores to control for
heterogeneity and ensure comparability across firms of varying scales and sectors. This
approach is grounded in the behavioural and Schumpeterian tradition of entrepreneurship,
which emphasises observable manifestations of innovation, risk taking, and growth-oriented
behaviour (Schumpeter, 1934; Davidsson and Wiklund, 2001; Gali et al., 2024).
A firm is deemed to demonstrate high growth when its compound annual growth rate
(CAGR) in sales or employment equals or exceeds 10% per annum for three consecutive years,
or when it consistently ranks within the top decile of its two-digit industry growth distribution.
This cut-off is consistent with European business statistics (EBS) regulation (OECD/Eurostat,
2025, 2018) for high growth enterprises and reflects the notion of entrepreneurial dynamism
and market expansion. Innovation intensity is operationalised as either R&D expenditure
exceeding 3% of annual turnover or at least one registered patent, design, or trademark within
the past five years, as verified through national or international registries. This indicator
reflects Schumpeterian innovation and aligns with the OECD/Eurostat Oslo Manual (2018)
definition of innovative activity. Entrepreneurial finance is established where a firm has
received venture capital, angel, or equity crowdfunding investment within the preceding five
years, verified through credible databases, such as Crunchbase, PitchBook, or public filings.
This dimension captures the firm’s capacity to mobilise high risk external capital, signifying
investor confidence in its growth and innovation potential (Kortum and Lerner, 2000;
Chemmanur and Fulghieri, 2014).
To reinforce replicability and prevent definitional drift, all thresholds are grounded in
internationally accepted standards and are cross validated against alternative calibrations, for
instance, 15% and 25% growth thresholds, or 2% and 4% R&D-to-sales ratios. This robustness
testing ensures that classification results remain stable under varying conditions and cut-off
values. Firms that meet at least two of the three dimensions are classified as entrepreneurial,
ensuring the inclusion of both early-stage ventures and established firms that sustain
entrepreneurial orientation through continuous innovation and strategic renewal. This
integrated operationalisation enhances conceptual clarity and empirical reliability while
safeguarding against selective inclusion by future scholars. It aligns with contemporary calls
for multidimensional, behaviourally anchored measures of entrepreneurship (Wiklund et al.,
2011; Zahra and Wright, 2011) and advances cumulative comparability across studies, thereby
contributing to a more standardised and replicable understanding of what constitutes an
entrepreneurial firm.
The sample breakdown by country is presented in Table 1, highlighting that Sweden
(18.04%), Germany (17.43%), France (13.13%), and Italy (8.53%) constitute the largest
contributors. These top four countries account for a substantial proportion of the final dataset,
reflecting their prominent role in AI and entrepreneurial activities within the EU.
3.2 Measures
3.2.1 Dependent variable. This study employs changes in revenue growth (ΔREV_GR),
calculated as the first difference of a firm’s revenue growth, as the dependent variable.
Revenue growth is widely regarded as an ideal measure of organisational performance and
entrepreneurial outcomes. Building on prior research (e.g., Czarnitzki et al., 2023; Groza et al.,
2021), it reflects management and organisational innovativeness, capturing the dual aspects of
driving revenue and gaining customer acceptance for firm innovations.
The conceptual underpinning aligns with the arguments of Katsikeas et al. (2016) and
Kohtam€aki et al. (2019), who emphasise that revenue growth encapsulates both the
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Table 1. Sample breakdown
IJEBR
Number Percentage
Country of firms Observations (%)
Austria 28 320 2.47
Belgium 52 503 3.89
Bulgaria 5 51 0.39
Croatia 6 56 0.43
Cyprus 5 50 0.39
Czechia 5 41 0.32
Denmark 66 482 3.72
Estonia 5 39 0.30
Finland 96 862 6.66
France 191 1,699 13.13
Germany 228 2,256 17.43
Greece 32 336 2.60
Hungary 2 24 0.19
Ireland 31 296 2.29
Italy 141 1,104 8.53
Latvia 3 30 0.23
Luxembourg 8 78 0.60
Malta 4 43 0.33
Netherlands 47 407 3.14
Poland 116 1,077 8.32
Portugal 8 90 0.70
Romania 8 86 0.66
Slovakia 3 29 0.22
Slovenia 4 36 0.28
Spain 63 612 4.73
Sweden 322 2,335 18.04
Total 1,479 12,942 100
Note(s): This table provides a detailed breakdown of the sample distribution across countries included in
the study
Source(s): Authors’ own work
commercial success of innovations and the entrepreneurial capacity of firms. As such, it serves
as a robust proxy for measuring entrepreneurship, particularly in innovation driven contexts.
Furthermore, adopting a longitudinal approach to revenue growth enhances the robustness of
causality inferences (Podsakoff et al., 2003).
Revenue growth alone adequately represents entrepreneurial outcomes because it
inherently captures the dynamic and multifaceted nature of entrepreneurship (Davidsson
and Wiklund, 2001; Stam et al., 2014). Entrepreneurial activities are fundamentally aimed at
creating economic value, which is most directly and objectively reflected in revenue metrics
(Covin and Slevin, 1989; Lumpkin and Dess, 1996). Unlike static measures such as
profitability or market share, revenue growth provides a continuous and scalable indicator of a
firm’s ability to identify opportunities, innovate, and respond to market demands effectively
(Davidsson et al., 2009; Wiklund and Shepherd, 2005). Additionally, revenue growth
embodies the cumulative effect of various entrepreneurial processes such as product
development, market expansion, customer acquisition, and strategic partnerships, thereby
offering a holistic view of entrepreneurial success (Delmar et al., 2003). This singular focus
also mitigates the complexities and potential biases associated with multi-dimensional
performance measures, ensuring clarity and comparability across different firms and contexts
(Brush et al., 2009; Rauch et al., 2009).
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

3.2.2 Independent variables. 3.2.2.1 Artificial intelligence (AI). One of the significant International
challenges in this study is identifying a suitable proxy for AI adoption and activity. The OECD Journal of
AI Policy Observatory offers an extensive range of AI related data, including AI news, Entrepreneurial
Behavior &
demographics, research, investments, jobs and skills, software development, search trends,
Research
education, knowledge flows, models and databases, and social media trends at both country
and industry levels. Given the focus and scope of this study, Venture Capital (VC) investments
in AI emerge as the most relevant proxy.
The dataset includes the total monetary value of investments (USD million) and the number
of VC investments in AI across countries and industries based on the data available in
PREQIN. Data is available for the years 2012–2023 for all 26 EU countries. This makes VC
investments an ideal indicator of AI adoption and its entrepreneurial impact, as it reflects both
the intensity of AI-related innovation and the confidence of investors in AI technologies.
This choice is justified for several reasons: VC investments signal market confidence and
highlight the intensity of innovation, as global AI focused VC funding has surged, exceeding
$100 billion in 2024, up from $55.6 billion in 2023 (Glaser and Yao, 2025). VC funding also
captures entrepreneurial activity since it is often directed toward startups and innovative firms
commercialising AI applications. Moreover, the availability of granular and comparable VC
data allows for robust and consistent cross-country and cross-industry analyses. This data
correlates with broader economic trends and reflects a significant shift toward AI driven
innovation, exemplified by the fact that AI companies received 42% of all U.S. venture capital
investments in 2024 (PYMNTS, 2024). Consequently, employing VC investment data as a
proxy enables a dynamic assessment of AI’s role in driving entrepreneurial activity and
revenue growth across diverse firms and contexts.
This study employs the log transformed VC investments in AI (AI_INV) as the main
independent variable. Following the approach of Felten et al. (2021), for each firm-year, the
total AI focused VC investment in that firm’s primary country, as reported by OECD’s AI
Observatory, is matched as a proxy for the firm’s AI investment intensity. This country-level
measure captures the broader AI investment environment influencing firm behaviour. The
logarithmic transformation mitigates issues of scale and variance, ensuring that the variable is
appropriately normalised for econometric analysis.
To examine the possibility of a U-shaped (quadratic) relationship between changes in
revenue growth (ΔREV_GR) and AI investments, the squared term of AI investments (AI_
2
INV ) is included in the model. The interaction of AI_INV with itself allows us to capture non-
linear effects and investigate whether changes in revenue growth initially decrease but later
increase as AI investment levels rise. This approach provides a meaningful understanding of
the complex dynamics between AI adoption and entrepreneurial outcomes, as hypothesised in
prior literature (e.g. Brynjolfsson et al., 2019; Agrawal et al., 2019).
By incorporating the quadratic term, the model accounts for potential diminishing or
increasing returns on revenue growth as AI investments scale, offering deeper insights into the
entrepreneurial implications of varying levels of AI activity across the EU.
3.2.2.2 Research and development (R&D). Following Czarnitzki et al. (2023), a dummy
variable is constructed for R&D based on the firm’s reported R&D expenditure. Specifically,
the variable is assigned a value of 1 if the firm reports R&D expenditure in a given year and
0 otherwise. This binary variable captures the presence of R&D activity, enabling us to
investigate its role in influencing changes in revenue growth (ΔREV_GR). In the study, a
dummy indicator for R&D activity is used to avoid issues arising from extreme variability in
R&D expenditures and because the distinction between firms engaging in formal innovation
activities versus not is most pertinent to the theoretical focus. In the sample, firms that report
R&D tend to be those with significant innovative commitments.
In addition to assessing the standalone impact of R&D expenditure on revenue growth, this
research also explores its interaction with AI investments (AI_INV) to understand how the
combination of R&D activities and AI adoption affects entrepreneurial outcomes. The
inclusion of this interaction term (R&D X AI_INV) allows us to test whether firms that
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

actively engage in R&D are better positioned to leverage AI technologies for driving revenue
IJEBR
growth. This is consistent with prior studies emphasising the complementary relationship
between technological adoption and innovation capabilities (e.g. Kohtam€aki et al., 2019;
Agrawal et al., 2019).
By incorporating both the standalone and interaction effects, this study provides a
comprehensive analysis of how R&D, as an indicator of organisational innovativeness,
strengthens the entrepreneurial benefits of AI investments. These findings are crucial for
understanding the conditions under which AI adoption yields the most significant performance
outcomes in firms across the EU.
3.2.3 Firm-specific and macroeconomic control variables. Based on prior research (e.g.
Czarnitzki et al., 2023; Groza et al., 2021; Brynjolfsson et al., 2019; Agrawal et al., 2019), this
study incorporates a comprehensive set of control variables to account for firm-specific and
macroeconomic factors influencing changes in revenue growth (ΔREV_GR). Firm-specific
controls include Firm Size (SIZE), measured as the natural logarithm of total assets, to capture
the scale of the firm’s operations and resource base. Firm Age (AGE) is defined as the number
of years since the firm’s establishment, reflecting organizational maturity and experience.
Gross Profit (ΔGPROFIT) represents the change in the natural logarithm of gross profit,
serving as an indicator of operational performance. Equity (ΔEQUITY) is captured as the
change in the natural logarithm of equity, highlighting shifts in financial stability and
shareholder value. Operating Expenses (ΔOPRT_EXP) are measured as the change in the
natural logarithm of operating expenses, representing adjustments in cost structure and
operational efficiency. Finally, Debt (ΔDEBT) is defined as the change in the natural logarithm
of total debt, accounting for variations in leverage and financial strategy.
This paper also includes macroeconomic controls to account for broader economic
influences. GDP Growth (ΔGDPG) measures changes in GDP growth, reflecting overall
economic performance and its potential impact on entrepreneurial activity. Inflation (ΔINF)
captures changes in inflation, providing insight into macroeconomic stability and purchasing
power. Institutional Quality (ΔIQ) is derived from the World Governance Indicators (WGI)
and includes six components: control of corruption, government effectiveness, political
stability and absence of violence/terrorism, regulatory quality, rule of law, and voice and
accountability, as defined by Kaufmann et al. (2010). These variables provide a
comprehensive framework to isolate the effects of AI investments on revenue growth while
accounting for firm level heterogeneity and macroeconomic conditions.
3.3 Estimation technique
To test the hypotheses, this study employs the Cameron et al. (2011) multi-way clustering
(CGM) estimation technique with heteroscedasticity-corrected, clustered robust standard
errors and robust interference. This approach effectively addresses two key econometric
challenges inherent in multi-dimensional panel data: within-group serial correlation and cross-
sectional dependence across firms, industries, countries, and time periods. By simultaneously
clustering along multiple dimensions, the CGM method provides more accurate standard
errors, enhancing the reliability of our estimates and mitigating potential biases associated
with unaccounted correlations (Petersen, 2008; Thompson, 2011). This technique is now
widely regarded as a best practice in empirical corporate finance and innovation studies,
making it particularly well-suited for the analysis of AI investment and entrepreneurial
performance across diverse contexts. This paper estimates panel regression models with
appropriate fixed effects (year, industry, and country fixed effects in most specifications, and
alternative specifications with firm fixed effects for robustness) and clustered standard errors.
This research further checked the robustness of the results by utilising an alternative proxy
for AI and dividing the sample based on firm size (large versus small/medium firms) to explore
potential heterogeneity in the effects. To mitigate endogeneity, this study implements a two-
stage least squares instrumental variable (2SLS-IV) regression using industry level AI
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

investment (i.e. total VC investment in AI in the firm’s industry across countries) as an International
instrument for firm level AI_INV. The assumption is that industry wide AI investment trends Journal of
drive individual firm AI adoption decisions but are not directly caused by any single firm’s Entrepreneurial
Behavior &
short term revenue changes. This robust methodology ensures the study produces reliable
Research
insights into the intricate dynamics between AI adoption and entrepreneurial outcomes,
particularly how R&D activities interact with AI adoption to influence these outcomes.
4. Results
4.1 Descriptive statistics and main results
Table 2 presents the descriptive statistics and definitions of the variables used in this study,
providing a clear overview of their characteristics and operationalisation. Table 3 reports the
pairwise correlations among the independent variables. The results in Table 3 indicate that
none of the variables exhibit high correlation coefficients, suggesting the absence of
multicollinearity concerns. This ensures the robustness of the regression analyses and the
reliability of the estimated relationships between AI investments, R&D, and changes in
revenue growth.
Table 4 reports the baseline results, examining the role of AI investments (AI_INV) and
their quadratic relationship with changes in revenue growth (ΔREV_GR), as well as the
impact of R&D and its interaction with AI on ΔREV_GR. Model 1 includes only firm-specific
control variables, while Model 2 incorporates both firm-specific and macroeconomic controls.
Model 3 applies year and firm fixed effects with industry clustering, Model 4 uses year and
industry fixed effects with firm clustering, and Models 5 and 6 employ year, country, and
industry fixed effects with firm clustering.
2
Models 1–5 evaluate the relationship between AI_INV, its quadratic term (AI_INV ), and
ΔREV_GR. The findings reveal an initially negative relationship between AI_INV and
2
ΔREV_GR, while AI_INV shows a positive and significant association at a 5% level of
significance. These findings suggest that early-stage AI investments can initially hinder
Table 2. Descriptive statistics
Variables Definition N Mean SD Min Max
REV_GR Year-over-year revenue growth (%) 12,942 12.681 32.94 73.659 192.457
AI_INV Log of Venture Capital (VC) Investment 12,431 4.606 2.293 2.188 8.58
in Artificial Intelligence (AI)
R&D A dummy variable is created where a 12,942 0.280 0.449 0 1
value of 1 is assigned if the firm reports
R&D expenditure, and 0 otherwise
SIZE Firm size measured by log of Total Assets 12,942 19.644 2.64 12.983 25.563
AGE Firm age measured by the number of 12,942 52.903 49.175 5 207
years since the firm was established
GPROFIT Log of Gross Profit 12,942 18.236 2.706 7.012 23.816
EQUITY Log of Total Equity 12,942 18.747 2.619 5.885 24.407
OPRT_EXP Log of Operating Expenses 12,942 18.136 2.412 10.996 23.383
DEBT Log of Total Debt 12,942 17.739 3.122 0.24 24.217
GDPG GDP Growth 12,942 1.582 3.111 8.868 8.931
INF Inflation 12,942 2.414 2.775 0.874 11.529
IQ Institutional Quality 12,942 1.223 0.443 0.189 1.861
Note(s): This table summarises the descriptive statistics and provides definitions for all variables used in the
analysis. N, SD, Min and Max refer to number of observations, standard deviation, minimum value and
maximum value respectively
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Table 3. Pairwise correlations
Variables 1 2 3 4 5 6 7 8 9 10
(1) AI_INV 1
(2) SIZE 0.109*** 1
(3) AGE 0.043*** 0.449*** 1
(4) ΔGPROFIT 0.031*** 0.038*** 0.061*** 1
(5) ΔEQUITY 0.014 0.042*** 0.065*** 0.168*** 1
(6) ΔOPRT_EXP 0.026*** 0.072*** 0.103*** 0.359*** 0.150*** 1
(7) ΔDEBT 0.017 0.004 0.036*** 0.046*** 0.015 0.102*** 1
(8) ΔGDPG 0.025*** 0.001 0.002 0.113*** 0.018 0.092*** 0.047*** 1
(9) ΔINF 0.352*** 0.046*** 0.028*** 0.037*** 0.004 0.034*** 0.014 0.122*** 1
(10) ΔIQ 0.035*** 0.020 0.006 0.045*** 0.007 0.056*** 0.006 0.243*** 0.033*** 1
Note(s): This table presents the pairwise correlation coefficients among the independent variables, highlighting the relationships and ensuring no significant multicollinearity
issues
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026
IJEBR

Table 4. Main results
International
Journal of
Dependent variable: ΔREV_GR
Entrepreneurial
Variables (1) (2) (3) (4) (5) (6)
Behavior &
Research
AI_INV 0.226 0.678 0.027 0.791* 0.679 0.187
(0.685) (0.668) (0.689) (0.479) (0.748) (0.579)
AI_INV^2 0.140** 0.139** 0.144** 0.128** 0.140**
(0.051) (0.054) (0.064) (0.051) (0.059)
R&D 1.212
(1.595)
AI_INV x R&D 0.504**
(0.267)
SIZE 0.155 0.156 5.606*** 0.290** 0.200* 0.132
(0.130) (0.131) (1.050) (0.117) (0.118) (0.128)
AGE 0.037*** 0.037*** 0.243 0.037*** 0.038*** 0.039***
(0.006) (0.006) (0.447) (0.005) (0.005) (0.005)
ΔGPROFIT 40.396*** 40.271*** 43.713*** 40.258*** 40.282*** 40.298***
(2.374) (2.401) (2.894) (2.690) (2.692) (2.692)
ΔEQUITY 1.090 1.170 4.204** 1.058 1.195 1.197
(1.408) (1.372) (1.334) (1.263) (1.267) (1.267)
ΔOPRT_EXP 0.990 0.831 3.732 1.009 0.833 0.826
(2.243) (2.252) (3.231) (2.485) (2.502) (2.504)
ΔDEBT 0.769 0.754 1.558** 0.740 0.743 0.756
(0.464) (0.461) (0.503) (0.557) (0.558) (0.557)
ΔGDPG 0.452** 0.423 0.462** 0.453** 0.447**
(0.202) (0.237) (0.195) (0.197) (0.197)
ΔINF 1.501*** 1.582*** 1.341*** 1.499*** 1.506***
(0.322) (0.343) (0.426) (0.457) (0.458)
ΔIQ 16.522* 11.619 14.660* 16.438* 18.868**
(7.602) (6.624) (7.858) (8.645) (8.470)
Constant 13.268*** 11.021*** 87.803** 13.518*** 11.982*** 11.878***
(3.231) (3.107) (36.514) (2.697) (3.158) (3.152)
Observations 10,751 10,751 10,751 10,751 10,751 10,751
Adjusted 0.219 0.221 0.219 0.221 0.220 0.220
R-squared
F Statistics 22.210*** 23.340*** 25.410*** 40.936*** 21.674*** 21.201***
Year fixed effect Yes Yes Yes Yes Yes Yes
Country fixed Yes Yes No No Yes Yes
effect
Firm fixed effect No No Yes No No No
Industry fixed No No No Yes Yes Yes
effect
Firm clustered Yes Yes No Yes Yes Yes
Industry Yes Yes Yes No No No
clustered
SE clustered Yes Yes Yes Yes Yes Yes
Note(s): The table shows the main results of the analysis. ΔREV_GR, AI_INV, R&D, SIZE, AGE, ΔGPROFIT,
ΔEQUITY, ΔOPRT_EXP, ΔDEBT, ΔGDPG, ΔINF, ΔIQ refer to the first difference of a firm’s revenue growth,
log of venture capital investment in artificial intelligence, a dummy variable is created where a value of 1 is
assigned if the firm reports R&D expenditure, and 0 otherwise, log of total assets, the number of years since the
firm was established, the first difference of a firm’s log gross profit, the first difference of a firm’s log total equity,
the first difference of a firm’s log operating expenses, the first difference of a firm’s log total debt, the first
difference of a country’s GDP growth, the first difference of a country’s inflation, the first difference of a
country’s institutional quality respectively. Robust and clustered standard errors are reported in parentheses.
Statistical significance is denoted by *, **, and *** for the 10%, 5%, and 1% levels, respectively. Data sources
include S&P Capital IQ, OECD AI Policy Observatory, World Development Indicators (WDI), and World
Governance Indicators (WGI)
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

revenue growth due to high upfront costs, integration challenges, or implementation
IJEBR
inefficiencies (Czarnitzki et al., 2023; Brynjolfsson et al., 2019; Agrawal et al., 2019).
However, as firms gain experience and optimise AI utilisation, the impact on revenue growth
turns positive, confirming a U-shaped relationship and supporting our first hypothesis (H1).
Economically, based on the regression estimates in Model 5, the turning point is observed at
a log AI investment value of approximately 2.425 ( 0.679/(2 3 0.140)), corresponding to
approximately USD 11.3 million in AI-related venture funding. This suggests that firms
investing below this level may face short term performance declines, while those exceeding
this threshold begin to realise the long-term benefits of AI adoption through efficiency gains,
innovation, and improved decision-making capabilities.
Model 6 (in Table 4) investigates the effects of AI, R&D, and their interaction (AI_
INV 3 R&D) on ΔREV_GR. The results reveal that AI investment alone has a positive but
statistically insignificant relationship with revenue growth, while R&D alone shows a
negative association. The negative coefficient for R&D may reflect that firms engaging in
R&D are incurring substantial upfront costs now for future payoffs, which can slightly slower
revenue growth in short-term. However, the interaction term (AI_INV 3 R&D) exhibits a
positive and statistically significant effect, indicating that the combination of AI adoption and
R&D activity produces complementary and mutually reinforcing effects on firm performance.
This result supports our second hypothesis (H2) and aligns with prior research highlighting
that integrating AI technologies with ongoing innovation activities amplifies productivity and
competitiveness (Czarnitzki et al., 2023; Kohtam€aki et al., 2019; Agrawal et al., 2019). These
findings emphasise that AI and R&D are most effective when deployed jointly, enabling firms
to transform technological capabilities into tangible entrepreneurial growth outcomes.
Among the control variables, the results align largely with expectations. Higher GDP
growth is positively associated with firm revenue growth, indicating that firms benefit from
favourable macroeconomic conditions. Inflation also shows a positive effect, possibly
reflecting firms’ ability to pass higher prices to consumers during expansionary periods. Firm-
specific controls show mixed results: firm age generally exhibits a positive relationship with
revenue growth, suggesting that more established firms may leverage accumulated experience
and market resilience, whereas firm size occasionally shows a negative association, consistent
with the notion that smaller firms often possess greater agility and higher growth potential.
4.2 Additional analysis and robustness test
This study conducted an additional analysis to explore differences in the relationship between
AI investments and revenue growth across large and small/medium-sized firms. Firm size was
determined based on total assets, with firms above the median value categorised as large and
those below categorised as small/medium-sized. Table 5 summarises the results, examining
2
the relationship between AI investments (AI_INV), their quadratic term (AI_INV ), and
changes in revenue growth (ΔREV_GR), as well as the effects of AI, R&D, and their
interaction (AI_INV 3 R&D) on ΔREV_GR across the two firm-size groups.
The findings indicate that for small/medium-sized firms, the results align closely with the
primary analysis, showing a U-shaped relationship between AI investments and revenue
growth, and a significant positive effect of the interaction between AI and R&D on revenue
growth. However, for large firms, no significant relationships were observed. This is due to the
fact that larger firms often face greater regulatory compliance pressures, bureaucratic
complexities, and multiple layers of decision-making, which can slow down the
implementation and responsiveness needed to fully capitalise on AI technologies. A recent
report by ANS (2025), based on a survey of 1,000 firms, further supports this view by
highlighting that larger organisations are subject to more stringent legal and compliance
requirements when adopting AI, which can hinder rapid integration and innovation. On the
other hand, this divergence suggests that smaller firms may experience more pronounced
benefits from AI adoption and R&D synergy, potentially due to their greater flexibility and
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Table 5. Sample split: Big firms vs Small & Medium firms
International
Journal of
Dependent variable: ΔREV_GR
Entrepreneurial
Medium and Medium and
Behavior &
small firms Big firms small firms Big firms
Research
Variables (7) (8) (9) (10)
AI_INV 1.711* 0.417 0.398 0.039
(0.974) (0.902) (0.261) (0.376)
AI_INV^2 0.224** 0.050
(0.091) (0.078)
R&D 0.418 2.612
(1.567) (1.790)
AI_INV x R&D 0.362* 0.642
(0.211) (0.388)
AGE 0.033*** 0.052*** 0.035*** 0.045***
(0.009) (0.009) (0.008) (0.009)
ΔGPROFIT 41.839*** 39.014*** 41.885*** 39.031***
(3.494) (3.713) (4.287) (4.285)
ΔEQUITY 0.664 1.607 0.586 1.463
(2.126) (1.389) (1.012) (1.978)
ΔOPRT_EXP 2.342 0.760 2.683 0.621
(4.001) (2.712) (4.417) (3.201)
ΔDEBT 0.814 0.767 0.764 0.770*
(0.814) (0.794) (0.484) (0.432)
ΔGDPG 0.535* 0.383* 0.521 0.356*
(0.290) (0.229) (0.324) (0.195)
ΔINF 2.071*** 0.925 1.549*** 1.039**
(0.648) (0.606) (0.409) (0.451)
ΔIQ 0.367 34.844*** 0.608 31.834***
(12.073) (12.628) (11.969) (10.020)
Constant 2.668 13.873*** 7.020*** 11.375***
(3.169) (2.786) (1.249) (1.403)
Observations 5,318 5,433 5,318 5,433
Adjusted R-squared 0.214 0.228 0.214 0.229
F Statistics 11.119*** 13.832*** 12814.679*** 4296.236***
Year fixed effect Yes Yes Yes Yes
Country fixed effect Yes Yes Yes Yes
Firm fixed effect No No No No
Industry fixed effect Yes Yes Yes Yes
Firm clustered Yes Yes Yes Yes
Industry clustered No No No No
SE clustered Yes Yes Yes Yes
Note(s): This table presents the comparative results of the relationship between AI investments and revenue
growth for large firms versus small/medium-sized firms. The analysis highlights the differences in how AI
investment impacts revenue growth across these two firm size categories. ΔREV_GR, AI_INV, R&D, SIZE,
AGE, ΔGPROFIT, ΔEQUITY, ΔOPRT_EXP, ΔDEBT, ΔGDPG, ΔINF, ΔIQ refer to the first difference of a
firm’s revenue growth, log of venture capital investment in artificial intelligence, a dummy variable is created
where a value of 1 is assigned if the firm reports R&D expenditure, and 0 otherwise, log of total assets, the
number of years since the firm was established, the first difference of a firm’s log gross profit, the first difference
of a firm’s log total equity, the first difference of a firm’s log operating expenses, the first difference of a firm’s
log total debt, the first difference of a country’s GDP growth, the first difference of a country’s inflation, the first
difference of a country’s institutional quality respectively. Robust and clustered standard errors are reported in
parentheses. Statistical significance is denoted by *, **, and *** for the 10%, 5%, and 1% levels, respectively.
Data sources include S&P Capital IQ, OECD AI Policy Observatory, World Development Indicators (WDI),
and World Governance Indicators (WGI)
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

ability to innovate compared to larger, more established organisations. These insights
IJEBR
highlight the importance of firm size in moderating the effects of AI and R&D on
entrepreneurial outcomes.
To ensure the reliability of the findings, a series of robustness checks is performed. Initially,
this study employed an alternative proxy for AI to assess the consistency of the results.
Subsequently, this research addressed potential endogeneity concerns by implementing a two-
stage least squares instrumental variable (2SLS-IV) regression using a valid instrument.
Finally, this paper employed various alternatives, such as profitability as an alternative proxy
for entrepreneurial outcome, AI intensity and patent filings as additional proxies for AI, and
R&D intensity as an alternative proxy for R&D.
For the alternative proxy, this study utilised the number of venture capital (VC) investments
in AI across countries, denoted as AI_NUM, which was derived from the OECD AI Policy
Observatory. Notably, the findings in Table 6 obtained using AI_NUM were consistent with
the primary results, reinforcing the robustness and credibility of our conclusions.
To address potential endogeneity concerns that might affect the relationship under study,
this paper adopted an instrumental variable (IV) approach following the methodology of
Czarnitzki et al. (2023). Industry level venture capital (VC) investment in AI across countries
was selected as the instrument for this analysis. This instrument captures variation in AI
investment at the industry level, mitigating firm level endogeneity issues. This choice is
theoretically justified and empirically robust for several reasons.
Industry level VC investment in AI captures sector specific technological trends,
competitive dynamics, and innovation pressures that are external to individual countries but
strongly influence country level AI investment decisions. In other words, when an industry at
the global or regional level (e.g., European IT or healthcare) sees a surge in VC funding for AI,
this often triggers increased AI related activity and investments in firms within the same
industry across different countries (Gompers and Lerner, 2001). These “technology push” and
“market pull” forces operate through knowledge spillovers, industry networks, and
competitive benchmarking (Jaffe et al., 1993). Therefore, the instrument is strongly
correlated with country level AI investment decisions, satisfying the relevance condition
(i.e. strong first-stage relationship).
At the same time, industry level VC investment in AI across countries is unlikely to be
directly influenced by revenue growth in any single country. Such investments are driven by
global venture capital trends, risk appetites, and macro-level perceptions of technological
potential (Kaplan and Lerner, 2016), rather than country specific firm revenue performance.
Moreover, they aggregate investment decisions across multiple industries and countries,
reducing potential feedback effects from individual country revenue changes back to the
instrument. This addresses concerns of reverse causality or direct dependence on the error term
in the second-stage regression, satisfying the exogeneity condition (Angrist and
Krueger, 2001).
The validity of the instrument was assessed through several diagnostic tests. A first-stage
F-test (results are available upon request) of the excluded instruments rejects the null
hypothesis, indicating that the instrument is relevant. Furthermore, the absolute F-value
exceeds the threshold of 10, alleviating concerns about weak instruments, as suggested by
Staiger and Stock (1997). The Hansen J-test for overidentifying restrictions does not reject the
null hypothesis in models (13) and (14) in Table 7, affirming the validity of the instrument.
With the instrument deemed appropriate based on the F-statistics and Hansen J-test, we re-
estimated the baseline regression using the IVapproach. The results of this analysis, consistent
with the initial findings, support the robustness of our conclusions regarding the relationship
between AI investments, R&D, and changes in revenue growth (ΔREV_GR). These
additional tests reinforce the reliability of the results and address potential concerns related to
endogeneity.
In addition to revenue growth, the study utilised return on equity (ΔPROFITABILITY) as a
proxy for profitability to assess the impact of AI on the firm’s profitability. We also considered
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Table 6. Alternative AI
International
Journal of
Dependent variable: ΔREV_GR
Entrepreneurial
Variables (11) (12)
Behavior &
Research
AI_NUM 2.171*** 0.066
(0.744) (0.439)
AI_NUM^2 0.411***
(0.124)
R&D 1.452
(1.397)
AI_NUM x R&D 0.806**
(0.376)
SIZE 0.291*** 0.257***
(0.068) (0.074)
AGE 0.037*** 0.035***
(0.006) (0.005)
ΔGPROFIT 40.091*** 40.115***
(3.279) (3.295)
ΔEQUITY 1.050 1.046
(1.089) (1.085)
ΔOPRT_EXP 0.926 0.938
(3.416) (3.425)
ΔDEBT 0.736* 0.745*
(0.384) (0.381)
ΔGDPG 0.448** 0.462**
(0.217) (0.217)
ΔINF 1.464*** 1.342***
(0.424) (0.373)
ΔIQ 14.818* 16.885*
(8.389) (8.192)
Constant 11.897*** 13.340***
(1.651) (1.612)
Observations 10,764 10,764
Adjusted R-squared 0.220 0.220
F Statistics 4755.192*** 12207.118***
Year fixed effect Yes Yes
Country fixed effect Yes Yes
Firm fixed effect No No
Industry fixed effect Yes Yes
Firm clustered Yes Yes
Industry clustered No No
SE clustered Yes Yes
Note(s): This table displays the key results obtained using an alternative proxy for AI investment, offering
additional validation for the primary findings. ΔREV_GR, AI_NUM, R&D, SIZE, AGE, ΔGPROFIT,
ΔEQUITY, ΔOPRT_EXP, ΔDEBT, ΔGDPG, ΔINF, ΔIQ refer to the first difference of a firm’s revenue growth,
log of number of venture capital investment in artificial intelligence, a dummy variable is created where a value
of 1 is assigned if the firm reports R&D expenditure, and 0 otherwise, log of total assets, the number of years
since the firm was established, the first difference of a firm’s log gross profit, the first difference of a firm’s log
total equity, the first difference of a firm’s log operating expenses, the first difference of a firm’s log total debt,
the first difference of a country’s GDP growth, the first difference of a country’s inflation, the first difference of a
country’s institutional quality respectively. Robust and clustered standard errors are reported in parentheses.
Statistical significance is denoted by *, **, and *** for the 10%, 5%, and 1% levels, respectively. Data sources
include S&P Capital IQ, OECD AI Policy Observatory, World Development Indicators (WDI), and World
Governance Indicators (WGI)
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Table 7. Endogeneity test: 2SLS-IV
IJEBR
Dependent variable: ΔREV_GR
Variables (13) (14)
AI_INV 0.679 0.187
(0.748) (0.579)
AI_INV^2 0.140**
(0.059)
R&D 1.212
(1.595)
AI_INV x R&D 0.504*
(0.267)
SIZE 0.200* 0.132
(0.118) (0.128)
AGE 0.038*** 0.039***
(0.005) (0.005)
ΔGPROFIT 40.282*** 40.298***
(2.692) (2.692)
ΔEQUITY 1.195 1.197
(1.267) (1.267)
ΔOPRT_EXP 0.833 0.826
(2.502) (2.504)
ΔDEBT 0.743 0.756
(0.558) (0.557)
ΔGDPG 0.453** 0.447**
(0.197) (0.197)
ΔINF 1.499*** 1.506***
(0.457) (0.458)
ΔIQ 16.438* 18.868**
(8.645) (8.470)
Observations 10,751 10,751
Centred R-squared 0.224 0.224
F Statistics 21.674*** 21.201***
Hansen J test statistics 1.340 1.741
Hansen J test (P value) 0.247 0.187
Year fixed effect Yes Yes
Country fixed effect Yes Yes
Firm fixed effect No No
Industry fixed effect Yes Yes
Firm clustered Yes Yes
Industry clustered No No
SE clustered Yes Yes
Note(s): This table presents the results addressing potential endogeneity concerns by employing valid
instruments, specifically industry-level venture capital (VC) investments in AI across countries. ΔREV_GR,
AI_INV, R&D, SIZE, AGE, ΔGPROFIT, ΔEQUITY, ΔOPRT_EXP, ΔDEBT, ΔGDPG, ΔINF, ΔIQ refer to the
first difference of a firm’s revenue growth, log of venture capital investment in artificial intelligence, a dummy
variable is created where a value of 1 is assigned if the firm reports R&D expenditure, and 0 otherwise, log of
total assets, the number of years since the firm was established, the first difference of a firm’s log gross profit, the
first difference of a firm’s log total equity, the first difference of a firm’s log operating expenses, the first
difference of a firm’s log total debt, the first difference of a country’s GDP growth, the first difference of a
country’s inflation, the first difference of a country’s institutional quality respectively. Robust and clustered
standard errors are reported in parentheses. Statistical significance is denoted by *, **, and *** for the 10%, 5%,
and 1% levels, respectively. Data sources include S&P Capital IQ, OECD AI Policy Observatory, World
Development Indicators (WDI), and World Governance Indicators (WGI)
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

AI intensity (AI_INT), which represents the venture capital investment divided by the total International
venture capital investment in AI as well as Patent filings, which represents the count of AI Journal of
related patent filings from 2012 to 2023 based on Google Patent data, as additional proxies for Entrepreneurial
Behavior &
AI to identify any notable differences from our initial findings. Furthermore, this paper
Research
incorporated R&D intensity (R&D_INT), which signifies the research and development
expenditure relative to total assets within a firm, as an alternative proxy to R&D to ascertain
any significant disparities. The outcomes of these alternative measures are presented in
Table 8. The results align with the initial findings and reinforce the robustness of the
conclusions regarding the correlation between AI investments, R&D, and fluctuations in
revenue growth (ΔREV_GR).
Table 8. Profitability, AI intensity, R&D intensity, and patent fillings
Dependent variable
ΔPROFITABILITY ΔREV_GR
Variables (15) (16) (17) (18) (19) (20)
AI_INV 1.361*** 0.651**
(0.219) (0.250)
AI_INV^2 0.101***
(0.021)
R&D 2.174 0.766
(1.363) (0.791)
AI_INV x R&D 0.489**
(0.208)
AI_INT 0.244*** 0.035 0.014
(0.044) (0.024) (0.046)
AI_INT^2 0.002***
(0.000)
AI_INT x R&D 0.057***
(0.012)
R&D_INT 2.452
(12.687)
AI_INT x R&D_INT 0.343***
(0.069)
PATENT 0.029***
(0.007)
Observations 10,735 10,735 10,764 10,764 3,026 5,090
Adjusted R-squared 0.150 0.150 0.210 0.210 0.224 0.266
Control Yes Yes Yes Yes Yes Yes
Country fixed effect Yes Yes Yes Yes Yes Yes
Firm clustered Yes Yes Yes Yes Yes Yes
Industry clustered Yes Yes Yes Yes Yes Yes
SE clustered Yes Yes Yes Yes Yes Yes
Note(s): ΔPROFITABILITY, ΔREV_GR, AI_INV, R&D, AI_INT, R&D_INT, PATENT refer to the first
difference of a firm’s return on equity, the first difference of a firm’s revenue growth, log of VC investment in AI,
a dummy variable is created where a value of 1 is assigned if the firm reports R&D expenditure, and 0 otherwise,
AI intensity which is the VC investment over the number of VC investment in AI, R&D intensity is the R&D
expenditure over total assets in a firm, number of patent filings related to AI based on data from Google patent
respectively. Robust and clustered standard errors are reported in parentheses. Statistical significance is denoted
by *, **, and *** for the 10%, 5%, and 1% levels, respectively. Data sources include S&P Capital IQ, OECD AI
Policy Observatory, Google Patent, World Development Indicators (WDI), and World Governance
Indicators (WGI)
Source(s): Authors’ own work
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

5. Discussion, contributions and policy implications
IJEBR
This study examined how AI investment influences entrepreneurial outcomes, proxied by
changes in firm revenue growth, among European firms. The results provide both theoretical
and practical insights.
The observed U-shaped relationship between AI investment and revenue growth
underscores the dual nature of technological adoption in entrepreneurship. Initially, firms
experience a decline in performance due to high implementation costs, skill mismatches, and
integration challenges, consistent with Brynjolfsson et al. (2019) and Agrawal et al. (2019).
However, over time, firms that persist through this learning phase realise substantial
performance gains as AI systems mature and organisational capabilities develop. This finding
aligns with Teece’s (2018) dynamic capability theory, emphasising that AI adoption is not a
static investment but a process of continuous adaptation and capability building.
The significant positive interaction between AI and R&D highlights the complementarity
of digital and innovation capabilities. Firms combining AI investments with R&D activities
are better equipped to generate, refine, and commercialise new products and services. The
negative short-term coefficient for R&D reflects delayed payoffs common to innovation
processes, but its interaction with AI reveals strong synergies that translate innovation
potential into market performance (Czarnitzki et al., 2023; Kohtam€aki et al., 2019). Among
the controls, GDP growth and inflation positively affected revenue growth, while firm age and
size showed mixed effects. Institutional quality (IQ) was insignificant, likely due to the EU’s
relatively harmonised regulatory and governance systems, suggesting limited cross-country
variation in short term outcomes.
Overall, the findings suggest that AI-driven entrepreneurship within the EU is most
effective in mature, innovation active firms with sufficient absorptive capacity to withstand
initial performance dips. Thus, the results apply primarily to established firms rather than
startups or micro-enterprises, which may face resource and capability constraints. Revenue
growth, as our proxy for entrepreneurial performance, captures firms’ ability to translate
innovation into financial outcomes, aligning with the broader entrepreneurship literature
(Davidsson and Wiklund, 2001).
5.1 Theoretical contributions
This study makes several important contributions to the RBT and the DCT. First, this research
extends RBT by exploring the threshold effects of AI investment in entrepreneurship. The
study identifies a U-shaped relationship between AI investments and entrepreneurial
outcomes, such as revenue growth, offering a new theoretical insight. This finding suggests
that the returns on AI investments are non-linear and may take time to appear, an aspect that has
been largely overlooked in existing research. Previous studies (Ameen et al., 2024; Giuggioli
and Pellegrini, 2023) have generally found a positive relationship between AI adoption and
entrepreneurial performance but have not considered the initial challenges firms face. High
implementation costs, inefficiencies in the workforce, integration difficulties and limited
technological knowledge can hinder performance during the early stages of AI adoption
(Dwivedi et al., 2021). As a result, firms may experience negative returns before achieving
long-term benefits.
By applying RBT to interpret this pattern, the findings suggest that AI investments initially
lack the characteristics of VRIN resources until firms build complementary assets and develop
learning routines that unlock their strategic potential (Barney, 1991). The U-shaped pattern
therefore provides empirical support for a non-linear relationship between resource
acquisition and value realisation, refining RBT by demonstrating that resources evolve
towards VRIN status over time through capability development (Makadok, 2001; Peteraf,
1993). This deepens theoretical understanding by showing that resource-based advantages are
not immediate results of ownership but are achieved through an adaptive process of
accumulation and recombination.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Second, this study highlights the synergistic relationship between AI investments and R&D International
efforts, thereby refining RBT’s assumptions about resource complementarity and bundling. Journal of
By examining how these two intangible assets interact to drive innovation and entrepreneurial Entrepreneurial
Behavior &
success, this research shows that the combination of AI and R&D generates complementary
Research
value that exceeds what either could achieve alone. This supports the principle of resource
orchestration, indicating that the strategic integration of AI and R&D enhances their combined
impact on performance (Sirmon et al., 2007, 2011). The findings therefore advance RBT by
illustrating that competitive advantage arises not only from possessing VRIN resources but
also from configuring and using them effectively within a coherent system of capabilities.
Third, the study extends RBT by considering the moderating role of firm size in the AI
R&D relationship. While previous studies (Ameen et al., 2024; Zhai and Liu, 2023) have
examined firm size as a factor influencing AI adoption, this research investigates how it shapes
the interaction between AI and R&D. The results indicate that smaller firms often benefit more
quickly from AI investments, whereas larger firms may experience diminishing returns unless
they adapt their integration strategies. This insight advances RBT by showing that the
effectiveness of strategic resources depends on organisational context, particularly firm size,
which affects how AI and R&D can be combined for competitive advantage. It also suggests
that the scalability of digital resources depends on managerial agility and the firm’s ability to
absorb and apply knowledge, extending RBT towards a more dynamic view of resource
effectiveness.
Fourth, this study contributes to the development of the DCT. While RBT explains what
resources firms possess, DCT focuses on how firms deploy, renew and reconfigure those
resources to remain competitive in changing environments. The findings suggest that AI
adoption functions not only as a valuable resource but also as a dynamic enabler that enhances
a firm’s ability to sense new opportunities, seize them through strategic action and reconfigure
existing processes to maintain long-term advantage. This explanation of the sensing, seizing
and reconfiguring mechanisms offers a clear illustration of how digital technologies operate as
sources of dynamic capability rather than as static inputs.
Finally, the study demonstrates that integrating AI with R&D strengthens long-term
adaptability and innovation, helping firms to maintain consistent performance in evolving
environments. This underlines the importance of integrated resource deployment in dynamic
conditions. While earlier research (Manfreda and Stemberger, 2019; Chen and Tian, 2022) has
emphasised that firms should integrate resources to build digital capabilities and achieve
successful transformation, this study identifies which specific resources, particularly AI and
R&D, are most important for driving entrepreneurial outcomes. By showing that AI R&D
integration improves both opportunity recognition and resource renewal, this study
strengthens DCT’s explanatory value in digital entrepreneurship.
This research bridges the conceptual divide between resource possession (RBT) and
resource orchestration (DCT), providing a more comprehensive theoretical framework for
understanding digital innovation and entrepreneurial performance. It refines RBT by showing
that the value of AI investment develops over time through learning and the accumulation of
complementary resources, and it extends DCT by explaining how AI supports the creation and
renewal of dynamic capabilities in practice. Together, these insights provide a stronger
theoretical foundation for understanding how digital technologies enable firms to achieve
sustainable entrepreneurial growth.
5.2 Managerial and policy implications
The findings offer actionable implications for both managers and policymakers.
For managers, the U-shaped relationship suggests that AI investments require strategic
patience and phased implementation. Firms should adopt staged investment strategies, starting
small, learning through pilot projects, and scaling gradually, to mitigate early inefficiencies
and manage initial costs. The positive AI and R&D interaction underscores the need for
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

alignment between digitalisation and innovation efforts. Managers should embed AI within
IJEBR
existing R&D pipelines to accelerate product development, enhance operational intelligence,
and sustain growth. Additionally, building complementary organisational capabilities, through
employee training, process redesign, and change management, is critical to realise AI’s full
potential.
For policymakers, the empirical evidence provides guidance on fostering sustainable AI-
driven entrepreneurship. The short-term negative effect of AI investments justifies transitional
policy support such as targeted tax incentives, innovation grants, and co-financing schemes,
particularly for small and medium-sized enterprises. Simultaneously, the strong
complementarity between AI and R&D calls for policies that promote joint investment
frameworks, for example linking digital adoption programmes with R&D funding under
Horizon Europe or Digital Europe.
Implementation must, however, consider cross-country heterogeneity in digital
infrastructure, institutional capacity, and firm demographics. Less digitally advanced
regions may require foundational support in connectivity and digital literacy before more
sophisticated AI adoption policies can take effect. Policymakers should also remain vigilant
about potential downsides, including over-subsidisation, uneven access to AI technologies,
and market concentration risks.
In terms of prioritisation, short-term policies should focus on easing adoption barriers
through financial incentives and workforce development, while long-term strategies should
emphasise innovation integration, ethical AI governance, and cross-border regulatory
harmonisation. Embedding these measures within existing EU frameworks such as Horizon
Europe, the Digital Europe Programme, and Cohesion Policy can enhance coordination,
ensuring that AI investments translate into broad-based entrepreneurial and economic growth.
6. Limitation and future research
This study provides novel insights into the relationship between AI investment and
entrepreneurial outcomes among European firms; however, several limitations remain, each of
which offers valuable opportunities for future research.
First, the sample composition focuses on R&D active, publicly listed firms across 26 EU
countries. These firms are typically more innovation oriented, resource rich, and structurally
capable of sustaining long-term technological investments. This strengthens internal validity
for analysing the AI–R&D interaction but may limit generalisability to smaller or less resource
intensive firms. Firms without dedicated R&D capacity might not be able to endure the short-
term challenges of AI adoption that underpin the observed U-shaped relationship, potentially
overstating the resilience of average entrepreneurial firms. Future research could extend this
analysis to include private, non-R&D, or early-stage firms to assess whether the AI
performance relationship differs across firm size, innovation capability, and funding structure.
Second, the measurement of AI investment relies on the logarithm of venture capital (VC)
investments in AI at the country levels, as provided by the OECD AI Policy Observatory.
While this proxy effectively captures aggregate AI ecosystem activity and investor
confidence, it does not fully reflect firm-specific AI adoption or capability levels.
Consequently, the results should be interpreted as reflecting broader AI environment effects
rather than direct firm-level behaviour. Future studies could employ firm-level AI indicators,
such as AI patents, AI employment intensity, or AI-related expenditure, to provide a more
granular understanding of adoption heterogeneity across firms and sectors.
Finally, the study’s time frame (2012–2023) captures a critical but still-evolving phase of
AI development. Major advances in deep learning emerged mid-decade, and generative AI and
large language models have only recently scaled. Thus, long-term effects may not yet be fully
observable. Although year fixed effects control for macro shocks such as COVID-19, future
research could extend the panel to include post-2023 data or employ dynamic structural
models to explore potential S-shaped or non-linear AI adoption effects.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

7. Conclusion International
In summary, AI investments can unlock significant entrepreneurial growth, but only through Journal of
an initially challenging learning period and in conjunction with innovation efforts and Entrepreneurial
Behavior &
supportive ecosystems. Entrepreneurs and policymakers must recognise that leveraging AI for
Research
sustained growth is a strategic, long-term endeavour, one that, when done right, can
substantially enhance competitive advantage and innovation in the economy.
References
Agrawal, A., Gans, J.S. and Goldfarb, A. (2019), “Exploring the impact of artificial intelligence:
prediction versus judgment”, Information Economics and Policy, Vol. 47, pp. 1-6, doi: 10.1016/
j.infoecopol.2019.05.001.
Ameen, N., Tarba, S., Cheah, J.H., Xia, S. and Sharma, G.D. (2024), “Coupling artificial intelligence
capability and strategic agility for enhanced product and service creativity”, British Journal of
Management, Vol. 35 No. 4, pp. 1916-1934, doi: 10.1111/1467-8551.12797.
Angrist, J.D. and Krueger, A.B. (2001), “Instrumental variables and the search for identification: from
supply and demand to natural experiments”, The Journal of Economic Perspectives, Vol. 15
No. 4, pp. 69-85, doi: 10.1257/jep.15.4.69.
ANS (2025), “AI readiness report 2025”, available at: https://www.ans.co.uk/data-ai/ai-readiness/
(accessed 1 October 2025).
Aziz, L.A.R. and Andriansyah, Y. (2023), “The role of artificial intelligence in modern banking: an
exploration of AI-driven approaches for enhanced fraud prevention, risk management, and
regulatory compliance”, Reviews of Contemporary Business Analytics, Vol. 6 No. 1, pp. 110-132.
Babina, T., Fedyk, A., He, A. and Hodson, J. (2024), “Artificial intelligence, firm growth, and product
innovation”, Journal of Financial Economics, Vol. 151, pp. 1-26, doi: 10.1016/j.jfineco.2023.103745.
Bahoo, S., Cucculelli, M. and Qamar, D. (2023), “Artificial intelligence and corporate innovation: a
review and research agenda”, Technological Forecasting and Social Change, Vol. 188, pp. 1-32,
doi: 10.1016/j.techfore.2022.122264.
Barney, J.B. (1991), “Firm resources and sustained competitive advantage”, Journal of Management,
Vol. 17 No. 1, pp. 99-120, doi: 10.1177/014920639101700108.
Barney, J.B. (2001), “Resource-based theories of competitive advantage: a ten-year retrospective on the
resource-based view”, Journal of Management, Vol. 27 No. 6, pp. 643-650, doi: 10.1177/
014920630102700602.
Brush, C.G., Ceru, D.J. and Blackburn, R. (2009), “Pathways to entrepreneurial growth: the influence
of management, marketing, and money”, Business Horizons, Vol. 52 No. 5, pp. 481-491, doi:
10.1016/j.bushor.2009.05.003.
Brynjolfsson, E. (1993), “The productivity paradox of information technology”, Communications of
the ACM, Vol. 36 No. 12, pp. 66-77, doi: 10.1145/163298.163309.
Brynjolfsson, E., Malone, T.W., Gurbaxani, V. and Kambil, A. (1994), “Does information technology
lead to smaller firms?”, Management Science, Vol. 40 No. 12, pp. 1628-1644, doi: 10.1287/
mnsc.40.12.1628.
Brynjolfsson, E., Hui, X. and Liu, M. (2019), “Does machine translation affect international trade?
Evidence from a large digital platform”, Management Science, Vol. 65 No. 12, pp. 5449-5460,
doi: 10.1287/mnsc.2019.3388.
Bughin, J., Seong, J., Manyika, J., Chui, M. and Joshi, R. (2018), “Notes from the AI Frontier:
modelling the impact of AI on the world economy”, McKinsey Global Institute.
Calabro, A., Vecchiarini, M., Gast, J., Campopiano, G., De Massis, A. and Kraus, S. (2019), “Innovation in
family firms: a systematic literature review and guidance for future research”, International Journal
of Management Reviews, Vol. 21 No. 3, pp. 317-355, doi: 10.1111/ijmr.12192.
Cameron, A.C., Gelbach, J.B. and Miller, D.L. (2011), “Robust inference with multiway clustering”,
Journal of Business and Economic Statistics, Vol. 29 No. 2, pp. 238-249, doi: 10.1198/
jbes.2010.07136.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Chalmers, D., MacKenzie, N.G. and Carter, S. (2021), “Artificial intelligence and entrepreneurship:
IJEBR
implications for venture creation in the fourth industrial revolution”, Entrepreneurship Theory
and Practice, Vol. 45 No. 5, pp. 1028-1053, doi: 10.1177/1042258720934581.
Chemmanur, T.J. and Fulghieri, P. (2014), “Entrepreneurial finance and innovation: an introduction and
agenda for future research”, Review of Financial Studies, Vol. 27 No. 1, pp. 1-19, doi: 10.1093/
rfs/hht063.
Chen, H. and Tian, Z. (2022), “Environmental uncertainty, resource orchestration and digital
transformation: a fuzzy-set QCA approach”, Journal of Business Research, Vol. 139,
pp. 184-193, doi: 10.1016/j.jbusres.2021.09.048.
Chen, K. and Zhang, S. (2023), “Impact of artificial intelligence on entrepreneurial activity: empirical
evidence from Chinese cities”, Technology Analysis and Strategic Management, Vol. 36 No. 12,
pp. 4701-4714, doi: 10.1080/09537325.2023.2268210.
Clark, E. (2024), “How small businesses and entrepreneurs can benefit from AI”, Forbes, available at:
https://www.forbes.com/sites/elijahclark/2023/11/30/how-small-businesses-and-entrepreneurs-
can-benefit-from-ai/ (accessed 27 December 2024).
Cohen, W.M. and Levinthal, D.A. (1990), “Absorptive capacity: a new perspective on learning and
innovation”, Administrative Science Quarterly, Vol. 35 No. 1, pp. 128-152, doi: 10.2307/
2393553.
Covin, J.G. and Slevin, D.P. (1989), “Strategic management of small firms in hostile and benign
environments”, Strategic Management Journal, Vol. 10 No. 1, pp. 75-87, doi: 10.1002/
smj.4250100107.
Cubric, M. (2020), “Drivers, barriers and social considerations for AI adoption in business and
management: a tertiary study”, Technology in Society, Vol. 62, pp. 1-12, doi: 10.1016/
j.techsoc.2020.101257.
Czarnitzki, D., Fern andez, G.P. and Rammer, C. (2023), “Artificial intelligence and firm-level
productivity”, Journal of Economic Behavior and Organization, Vol. 211, pp. 188-205, doi:
10.1016/j.jebo.2023.05.008.
Davenport, T.H. and Ronanki, R. (2018), “Artificial intelligence for the real world”, Harvard Business
Review, Vol. 96 No. 1, pp. 108-116.
Davidsson, P. and Wiklund, J. (2001), “Levels of analysis in entrepreneurship research: current
research practice and suggestions for the future”, Entrepreneurship Theory and Practice, Vol. 25
No. 4, pp. 81-100, doi: 10.1177/104225870102500406.
Davidsson, P., Steffens, P. and Fitzsimmons, J. (2009), “Growing profitable or growing from profits:
putting the horse in front of the cart?”, Journal of Business Venturing, Vol. 24 No. 4,
pp. 388-406, doi: 10.1016/j.jbusvent.2008.04.003.
Delmar, F., Davidsson, P. and Gartner, W.B. (2003), “Arriving at the high-growth firm”, Journal of
Business Venturing, Vol. 18 No. 2, pp. 189-216, doi: 10.1016/s0883-9026(02)00080-0.
Dicuonzo, G., Donofrio, F., Fusco, A. and Shini, M. (2023), “Healthcare system: moving forward with
artificial intelligence”, Technovation, Vol. 120, pp. 1-9, doi: 10.1016/
j.technovation.2022.102510.
Dubey, R., Gunasekaran, A., Childe, S.J., Fosso Wamba, S., Roubaud, D. and Foropon, C. (2021),
“Empirical investigation of data analytics capability and organizational flexibility as
complements to supply chain resilience”, International Journal of Production Research, Vol. 59
No. 1, pp. 110-128, doi: 10.1080/00207543.2019.1582820.
Dwivedi, Y.K., Hughes, L., Ismagilova, E., Aarts, G., Coombs, C., Crick, T., Duan, Y., Dwivedi, R.,
Edwards, J., Eirug, A., Galanos, V., Ilavarasan, P.V., Janssen, M., Jones, P., Kar, A.K., Kizgin,
H., Kronemann, B., Lal, B., Lucini, B., Medaglia, R., Le Meunier-FitzHugh, K., Le Meunier-
FitzHugh, L.C., Misra, S., Mogaji, E., Sharma, S.K., Singh, J.B., Raghavan, V., Raman, R.,
Rana, N.P., Samothrakis, S., Spencer, J., Tamilmani, K., Tubadji, A., Walton, P. and Williams,
M.D. (2021), “Artificial intelligence (AI): multidisciplinary perspectives on emerging
challenges, opportunities, and agenda for research, practice and policy”, International Journal of
Information Management, Vol. 57, pp. 1-47, doi: 10.1016/j.ijinfomgt.2019.08.002.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Easterby-Smith, M. and Prieto, I.M. (2008), “Dynamic capabilities and knowledge management: an
International
integrative role for learning?”, British Journal of Management, Vol. 19 No. 3, pp. 235-249, doi:
Journal of
10.1111/j.1467-8551.2007.00543.x.
Entrepreneurial
Behavior &
Felten, E., Raj, M. and Seamans, R. (2021), “Occupational, industry, and geographic exposure to
artificial intelligence: a novel dataset and its potential uses”, Strategic Management Journal, Research
Vol. 42 No. 12, pp. 2195-2217, doi: 10.1002/smj.3286.
Fu€ller, J., Hutter, K., Wahl, J., Bilgram, V. and Tekic, Z. (2022), “How AI revolutionizes innovation
management: perceptions and implementation preferences of AI-based innovators”,
Technological Forecasting and Social Change, Vol. 178, pp. 1-22, doi: 10.1016/
j.techfore.2022.121598.
Gali, N., Hughes, M., Morgan, R.E. and Wang, C.L. (2024), “Entrepreneurial entropy: a resource
exhaustion theory of firm failure from entrepreneurial orientation”, Entrepreneurship Theory
and Practice, Vol. 48 No. 1, pp. 141-170, doi: 10.1177/10422587231151957.
Ghazwani, S., van Esch, P., Cui, Y.G. and Gala, P. (2022), “Artificial intelligence, financial anxiety and
cashier-less checkouts: a Saudi Arabian perspective”, International Journal of Bank Marketing,
Vol. 40 No. 6, pp. 1200-1216, doi: 10.1108/ijbm-09-2021-0444.
Giuggioli, G. and Pellegrini, M.M. (2023), “Artificial intelligence as an enabler for entrepreneurs: a
systematic literature review and an agenda for future research”, International Journal of
Entrepreneurial Behavior and Research, Vol. 29 No. 4, pp. 816-837, doi: 10.1108/ijebr-05-
2021-0426.
Glaser, J.D. and Yao, K.M. (2025), “The state of the funding market for AI companies”, available at:
https://natlawreview.com/article/state-funding-market-ai-companies-2024-2025-outlook
(accessed 10 October 2025).
Gompers, P. and Lerner, J. (2001), “The venture capital revolution”, The Journal of Economic
Perspectives, Vol. 15 No. 2, pp. 145-168, doi: 10.1257/jep.15.2.145.
Groza, M.D., Zmich, L.J. and Rajabi, R. (2021), “Organizational innovativeness and firm performance:
does sales management matter?”, Industrial Marketing Management, Vol. 97, pp. 10-20.
Iansiti, M. and Lakhani, K.R. (2020), Competing in the Age of AI: Strategy and Leadership when
Algorithms and Networks Run the World, Harvard Business Review Press, Boston, MA.
Jaffe, A.B., Trajtenberg, M. and Henderson, R. (1993), “Geographic localization of knowledge
spillovers as evidenced by patent citations”, Quarterly Journal of Economics, Vol. 108 No. 3,
pp. 577-598, doi: 10.2307/2118401.
Jen, H.C. and Ju, L.C. (2005), “Exploration for the relationship between innovation, IT and
performance”, Journal of Intellectual Capital, Vol. 6 No. 2, pp. 237-252.
Jia, N., Luo, X., Fang, Z. and Liao, C. (2024), “When and how artificial intelligence augments
employee creativity”, Academy of Management Journal, Vol. 67 No. 1, pp. 5-32, doi: 10.5465/
amj.2022.0426.
Kang, K.H. and Kang, J. (2009), “How do firms source external knowledge for innovation? Analysing
effects of different knowledge sourcing methods”, International Journal of Innovation
Management, Vol. 13 No. 1, pp. 1-17, doi: 10.1142/s1363919609002194.
Kaplan, S.N. and Lerner, J. (2016), “Venture capital data: opportunities and challenges”, in John
Haltiwanger, et al. (Eds), Measuring Entrepreneurial Businesses: Current Knowledge and
Challenges, University of Chicago Press, Chicago, IL, pp. 413-431.
Kar, S., Kar, A.K. and Gupta, M.P. (2021), “Modeling drivers and barriers of artificial intelligence
adoption: insights from a strategic management perspective”, Intelligent Systems in Accounting,
Finance and Management, Vol. 28 No. 4, pp. 217-238, doi: 10.1002/isaf.1503.
Katsikeas, C.S., Morgan, N.A., Leonidou, L.C. and Hult, G.T.M. (2016), “Assessing performance
outcomes in marketing”, Journal of Marketing, Vol. 80 No. 2, pp. 1-20, doi: 10.1509/
jm.15.0287.
Kaufmann, D., Kraay, A. and Mastruzzi, M. (2010), “The worldwide governance indicators:
methodology and analytical issues”, World Bank Policy Research Working Paper No. 5430,
World Bank.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Kohtam€aki, M., Heimonen, J. and Parida, V. (2019), “The nonlinear relationship between
IJEBR
entrepreneurial orientation and sales growth: the moderating effects of slack resources and
absorptive capacity”, Journal of Business Research, Vol. 100, pp. 100-110, doi: 10.1016/
j.jbusres.2019.03.018.
Kortum, S. and Lerner, J. (2000), “Assessing the contribution of venture capital to innovation”, The
RAND Journal of Economics, Vol. 31 No. 4, pp. 674-692, doi: 10.2307/2696354.
Kraus, S., Schiavone, F., Pluzhnikova, A. and Invernizzi, A.C. (2021), “Digital transformation in
healthcare”, Journal of Business Research, Vol. 123, pp. 557-567, doi: 10.1016/
j.jbusres.2020.10.030.
Loveman, G.W. (1994), “An assessment of the productivity impact of information technologies”, in
Allen, T.J. and Scott Morton, M.S. (Eds), Information Technology and the Corporation of the
1990s, Oxford University Press, New York, NY, pp. 84-110.
Lui, A.K., Lee, M.C. and Ngai, E.W. (2022), “Impact of artificial intelligence investment on firm
value”, Annals of Operations Research, Vol. 308 No. 1, pp. 373-388, doi: 10.1007/s10479-020-
03862-8.
Lumpkin, G.T. and Dess, G.G. (1996), “Clarifying the entrepreneurial orientation construct and linking
it to performance”, Academy of Management Review, Vol. 21 No. 1, pp. 135-172, doi: 10.5465/
amr.1996.9602161568.
Madhani, P.M. (2010), “Resource-based view (RBV) of competitive advantage: an overview”, in
Madhani, P. (Ed.), Resource-Based View: Concepts and Practices, pp. 3-22.
Makadok, R. (2001), “Toward a synthesis of the resource-based and dynamic-capability views of
rent creation”, Strategic Management Journal, Vol. 22 No. 5, pp. 387-401, doi: 10.1002/
smj.158.
Manfreda, A. and Indihar Stemberger, M. (2019), “Establishing a partnership between top and IT
managers: a necessity in an era of digital transformation”, Information Technology and People,
Vol. 32 No. 4, pp. 948-972, doi: 10.1108/itp-01-2017-0001.
Mariani, M., Machado, I., Magrelli, V. and Dwivedi, Y.K. (2023), “Types of innovation and artificial
intelligence: a systematic quantitative literature review and research agenda”, Journal of
Business Research, Vol. 155, 113364.
Nambisan, S., Wright, M. and Feldman, M. (2019), “The digital transformation of innovation and
entrepreneurship: progress, challenges and key themes”, Research Policy, Vol. 48 No. 8, pp. 1-9,
doi: 10.1016/j.respol.2019.03.018.
Naveed, H. (2022), “Chatbots and virtual assistants: the future of customer service”, Computer Science
Bulletin, Vol. 5 No. 1, pp. 15-36.
OECD (2025), “The effects of generative AI on productivity, innovation and entrepreneurship”, OECD
Publishing, Paris.
OECD/Eurostat (2018), Oslo manual 2018: guidelines for collecting, reporting and using data on
innovation, OECD Publishing, Paris.
OECD/Eurostat (2025), “High-growth enterprises — statistics”, Statistics Explained, available at:
https://ec.europa.eu/eurostat/statistics-explained/index.php?title5High-growth_enterprises_-_
statistics#Footnotes (accessed 12 October 2025).
Peteraf, M.A. (1993), “The cornerstones of competitive advantage: a resource-based view”, Strategic
Management Journal, Vol. 14 No. 3, pp. 179-191, doi: 10.1002/smj.4250140303.
Petersen, M.A. (2008), “Estimating standard errors in finance panel data sets: comparing approaches”,
Review of Financial Studies, Vol. 22 No. 1, pp. 435-480, doi: 10.1093/rfs/hhn053.
Pham, P., Zhang, H., Gao, W. and Zhu, X. (2024), “Determinants and performance outcomes of
artificial intelligence adoption: evidence from US hospitals”, Journal of Business Research,
Vol. 172, pp. 1-18, doi: 10.1016/j.jbusres.2023.114402.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y. and Podsakoff, N.P. (2003), “Common method biases in
behavioral research: a critical review of the literature and recommended remedies”, Journal of
Applied Psychology, Vol. 88 No. 5, pp. 879-903, doi: 10.1037/0021-9010.88.5.879.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Priem, R.L. and Butler, J.E. (2001), “Is the resource-based “view” a useful perspective for strategic
International
management research?”, Academy of Management Review, Vol. 26 No. 1, pp. 22-40, doi:
Journal of
10.2307/259392.
Entrepreneurial
Behavior &
Przegalinska, A., Triantoro, T., Kovbasiuk, A., Ciechanowski, L., Freeman, R.B. and Sowa, K. (2025),
“Collaborative AI in the workplace: enhancing organizational performance through resource- Research
based and task-technology fit perspectives”, International Journal of Information Management,
Vol. 81, pp. 1-23, doi: 10.1016/j.ijinfomgt.2024.102853.
PwC (2015), “Maximizing the value from technology investment: spending smart instead of just
spending big”, Strategy Report.
PYMNTS (2024), “AI companies receive 42% of US venture capital investment”, available at:
PYMNTS.com, https://www.pymnts.com/artificial-intelligence-2/2024/ai-companies-receive-
42-of-us-venture-capital-investment/ (accessed 10 October 2025).
Rauch, A., Wiklund, J., Lumpkin, G.T. and Frese, M. (2009), “Entrepreneurial orientation and
business performance: an assessment of past research and suggestions for the future”,
Entrepreneurship Theory and Practice, Vol. 33 No. 3, pp. 761-787, doi: 10.1111/j.1540-
6520.2009.00308.x.
Rayner, B. (1995), “Management matters”, Computerworld Premier 100, pp. 7-10.
Santos, B. and Sussman, L. (2000), “Improving the return on IT investment: the productivity paradox”,
International Journal of Information Management, Vol. 20 No. 6, pp. 429-440, doi: 10.1016/
s0268-4012(00)00037-2.
Sarasvathy, S.D. (2001), “Causation and effectuation: toward a theoretical shift from economic
inevitability to entrepreneurial contingency”, Academy of Management Review, Vol. 26 No. 2,
pp. 243-263, doi: 10.2307/259121.
Schumpeter, J.A. (1934), The Theory of economic Development: An Inquiry Into Profits, Capital,
Credit, Interest, and the Business Cycle, Harvard University Press, Cambridge, MA.
Schumpeter, J.A. and Swedberg, R. (2021), The Theory of Economic Development, Routledge, London.
Sirmon, D.G., Hitt, M.A. and Ireland, R.D. (2007), “Managing firm resources in dynamic
environments to create value: looking inside the black box”, Academy of Management Review,
Vol. 32 No. 1, pp. 273-292, doi: 10.5465/amr.2007.23466005.
Sirmon, D.G., Hitt, M.A., Ireland, R.D. and Gilbert, B.A. (2011), “Resource orchestration to create
competitive advantage: breadth, depth, and life cycle effects”, Journal of Management, Vol. 37
No. 5, pp. 1390-1412, doi: 10.1177/0149206310385695.
Staiger, D. and Stock, J.H. (1997), “Instrumental variables regression with weak instruments”,
Econometrica, Vol. 65 No. 3, pp. 557-586, doi: 10.2307/2171753.
Stam, W., Arzlanian, S. and Elfring, T. (2014), “Social capital of entrepreneurs and small firm
performance: a meta-analysis of contextual and methodological moderators”, Journal of
Business Venturing, Vol. 29 No. 1, pp. 152-173, doi: 10.1016/j.jbusvent.2013.01.002.
Statista (2025), Artificial Intelligence (AI) Worldwide – Statistics & Facts, available at: https://
www.statista.com/topics/3104/
Suhyeon, P., Ruas, R.L. and Ouros, L.O. (2023), “Insights into the relationship between open
innovation and external knowledge acquisition in technology-based South Korean SMEs”,
International Journal of Innovation, Vol. 11 No. 2, pp. 9-22.
Sun, T.Q. and Medaglia, R. (2019), “Mapping the challenges of artificial intelligence in the public
sector: evidence from public healthcare”, Government Information Quarterly, Vol. 36 No. 2,
pp. 368-383, doi: 10.1016/j.giq.2018.09.008.
Teece, D.J. (2018), “Business models and dynamic capabilities”, Long Range Planning, Vol. 51 No. 1,
pp. 40-49, doi: 10.1016/j.lrp.2017.06.007.
Teece, D.J., Pisano, G. and Shuen, A. (1997), “Dynamic capabilities and strategic management”,
Strategic Management Journal, Vol. 18 No. 7, pp. 509-533.
Thompson, S.B. (2011), “Simple formulas for standard errors that cluster by both firm and time”,
Journal of Financial Economics, Vol. 99 No. 1, pp. 1-10, doi: 10.1016/j.jfineco.2010.08.016.
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026

Townsend, D.M., Hunt, R.A., McMullen, J.S. and Sarasvathy, S.D. (2025a), “Artificial intelligence and
IJEBR
the future of entrepreneurial decision-making”, Entrepreneurship Theory and Practice, Vol. 49
No. 2, pp. 365-387.
Townsend, D.M., Hunt, R.A., Rady, J., Manocha, P. and Jin, J.H. (2025b), “Are the futures
computable? Knightian uncertainty and artificial intelligence”, Academy of Management Review,
Vol. 50 No. 2, pp. 415-440, doi: 10.5465/amr.2022.0237.
Wiklund, J. and Shepherd, D. (2005), “Entrepreneurial orientation and small business performance: a
configurational approach”, Journal of Business Venturing, Vol. 20 No. 1, pp. 71-91, doi:
10.1016/j.jbusvent.2004.01.001.
Wiklund, J., Davidsson, P., Audretsch, D.B. and Karlsson, C. (2011), “The future of entrepreneurship
research”, Entrepreneurship Theory and Practice, Vol. 35 No. 1, pp. 1-9, doi: 10.1111/j.1540-
6520.2010.00420.x.
Wilson, H.J. and Daugherty, P.R. (2018), “Collaborative intelligence: humans and AI are joining
forces”, Harvard Business Review, Vol. 96 No. 4, pp. 114-123.
Wong, D.T. and Ngai, E.W. (2025), “Impact of artificial intelligence (AI) on operational performance:
the role of dynamic capabilities”, Information and Management, Vol. 62 No. 6, pp. 1-31, doi:
10.1016/j.im.2025.104162.
Zahra, S.A. and Wright, M. (2011), “Entrepreneurship’s next act”, Academy of Management
Perspectives, Vol. 25 No. 4, pp. 67-83, doi: 10.5465/amp.2010.0149.
Zhai, S. and Liu, Z. (2023), “Artificial intelligence technology innovation and firm productivity:
evidence from China”, Finance Research Letters, Vol. 58, pp. 1-9, doi: 10.1016/
j.frl.2023.104437.
Corresponding author
Hasanul Banna can be contacted at: b.banna@mmu.ac.uk
For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com
Downloaded from http://www.emerald.com/ijebr/article-pdf/doi/10.1108/IJEBR-12-2024-1489/11126093/ijebr-12-2024-1489en.pdf by University of Salford user on 15 June 2026