Artificial intelligence (AI) is increasingly embedded in safety-critical industries to support risk evaluation and decision-making. While AI holds considerable promise for reducing human error and enhancing efficiency, ethical concerns arise when AI advice leads individuals to relax safety standards and underestimate risks. This study investigates how the accuracy of AI advice shapes human safety evaluations in high-risk industrial contexts. A 2 (AI advice: correct vs. erroneous, between-subjects) & times; 2 (evaluation phase: independent phase vs. collaborative phase, within-subjects) mixed factorial design was implemented, in which 82 participants first independently evaluated the safety of high-risk industrial projects and subsequently re-evaluated them in collaboration with AI. Results demonstrated that erroneous AI advice was associated with significantly higher safety ratings of high-risk industrial contexts and a greater likelihood that participants reversed correct judgments into erroneous approvals. Moral disengagement emerged as a key mediator. When exposed to erroneous "safe" advice, participants exhibited stronger tendencies toward moral disengagement, which were in turn consistent with participants' rationalization of unsafe decisions and relaxation of safety thresholds. Moreover, safety evaluation experience moderated this effect, such that experienced individuals showed greater critical supervision and reduced susceptibility to erroneous AI advice. Together, these findings illuminate the cognitiveethical mechanism through which AI misguides human judgment in safety-critical contexts, revealing that while AI may reduce human error, it may simultaneously amplify automation bias and ethical vulnerability in collaborative decision-making.

Cultural heritage entrepreneurship (CHE) often requires entrepreneurs to reconcile authenticity claims with commercial imperatives. This challenge is especially pronounced in grassroots heritage revivals where heritage meanings are contested and cultural authority is uneven, yet CHE research has largely centered on ventures operating within authorized heritage discourse (AHD). Grounded in institutional theory, we draw on institutional logics and institutional work perspective to examine Hanfu entrepreneurship in China, a community-driven revival that remains largely outside the state-led heritage system while supporting a fast-growing consumer market. Based on a qualitative multiple-case study of seven Hanfu enterprises using interviews, ethnographic observation, and archival and digital trace data, we identify a community-generated Hanfu cultural inheritance logic and a commercial logic that operate alongside a dominant AHD cultural inheritance logic, as well as professionalism-oriented legitimation practices that render grassroots claims more publicly defensible without formal endorsement. We further show that entrepreneurs’ hybrid role identities channel distinct institutional work repertoires. Businessman-inheritors translate inheritance into market-legible offerings, reframe authenticity toward lived cultural resonance, and build de facto legitimacy through diffusion and public engagement. Artisan-inheritors mobilize evidence-based reconstruction, frame historical fidelity as verifiable expertise, and pursue boundary-spanning collaborations with authoritative cultural arenas. This study advances CHE by theorizing grassroots entrepreneurship as heritage-making under contested classification and extends institutional theory by linking logics, identity configurations, and identity-shaped institutional work in fields marked by uneven cultural authority.

Purpose This study aims to investigate how geopolitical tensions shape firms' innovation strategies. The effects on both incremental and radical innovation were examined in the context of the USA-China trade war. The moderating role of performance relative to aspirations (PRA) was further explored.Design/methodology/approach This study used a multi-source panel data set of 339 Chinese multinational enterprises with subsidiaries in the USA. From 2012 to 2022, and tested this study's hypotheses using a staggered difference-in-differences approach with firm and year fixed effects.Findings Firms in industries exposed to the US Entity List are more likely to increase radical innovation while decreasing incremental innovation. The negative effect on incremental innovation is amplified when performance falls below aspirations, while the positive effect on radical innovation is strengthened when performance exceeds aspirations.Originality/value This study contributes to the intersection of geopolitical and firm innovation strategy. It reveals how geopolitical tensions, along with firms' PRA, interact to shape their innovation strategies, offering new insights into corporate innovation under geopolitical tensions.

Purpose Although learning from innovation failures (LFIF) is essential for sustaining innovation and competitive advantage, individuals often struggle to overcome the cognitive and motivational barriers that impede deep reflection on failure. Drawing on reflection theory, this study conceptualizes AI as a reflection partner that reshapes how individuals interpret and learn from failures. On this basis, the study examines how human–AI collaboration facilitates LFIF through critical reflection and under what organizational contextual conditions this effect holds. Design/methodology/approach This study comprised a three-wave longitudinal survey of 221 product managers across diverse regions and industries in China. This temporal design mitigated common method bias and captured the sequencing of the proposed mediation process. Findings The results reveal that human–AI collaboration fosters LFIF indirectly through critical reflection. Specifically, human–AI collaboration stimulates alternative interpretations and challenges assumptions, thereby deepening reflection and enabling LFIF. Moreover, an organizational proactive climate strengthens the effect of human–AI collaboration on LFIF via critical reflection, while an organizational competitive climate weakens this indirect effect. Originality/value This study extends research on human–AI collaboration from routine to non-routine learning, theorizing AI as a reflection partner that reshapes higher-order cognition. Moreover, it identifies critical reflection as a central cognitive mechanism linking human–AI collaboration to LFIF, providing a new theoretical perspective for understanding LFIF in the AI era. By bridging technological, cognitive and contextual perspectives, this study elucidates when and how human–AI collaboration enables effective LFIF.

Organizations facing unfavorable performance feedback often engage in problemistic search. However, prior research has primarily focused on value creation, paying limited attention to value appropriation as a strategic response. We address this gap by examining how firms respond to different types of performance shortfalls across both value creation and value appropriation. We propose that historical performance shortfalls are processed through a diagnostic mechanism that primarily triggers value creation. In contrast, peer performance shortfalls are processed through an external attributional mechanism that primarily triggers value appropriation. Using panel data from 2,523 Chinese manufacturing firms covering 18,998 firm-year observations, we find strong support for these arguments. Extending previous studies that have predominantly focused on internal push forces, we further investigate how customer dynamics, as an external pull force, influence firms’ strategic responses to performance shortfalls. Our findings show that customer concentration dampens, whereas customer stability strengthens these relationships.

Emerging technologies are transforming how entrepreneurs identify, pursue, and develop new business opportunities. Establishing open-source software (OSS) communities has become a key competitive strategy for digital startups. However, firm-led OSS communities face persistent tension between core appropriability and peripheral generativity, which makes contribution outcomes uncertain. Framing a startup's initial public offering (IPO) as a boundary-reconfiguring event, we use a linked GitHub and Crunchbase panel dataset (2008-2021) to examine how an IPO reshapes both the overall level of contributions and the internal-external composition and destinations of contributions in the firm-led OSS community. We also examine heterogeneity by conditioning on preIPO boundary baselines and compare governance formalization with IPO signaling to clarify the conditions and pathways through which IPOs reconfigure OSS contributions. Our study extends the organizational boundary perspective to the open-innovation context and illustrates the dialectical relationship between community leaders and complementors. It also offers advice to startups and maintainers on when to tighten core governance and when to widen peripheral interfaces so that commercialization does not erode ecosystem generativity.

Sustainable food systems are now negotiated within a polycrisis defined by climate and biodiversity breakdown, persistent malnutrition, geopolitical fragmentation, and accelerating technological change. However, the prevailing innovation discourse in agri-food systems remains constrained by two recurring simplifications: an economistic framing of ‘innovation’ as productivity enhancement, and a technocratic framing of ‘digital transformation’ as inherently progressive. In this editorial, we position the International Journal of Innovation Studies special issue on Digital Transformation, Stakeholder Engagement, and Innovation in Sustainable Food Systems as an opportunity to re-politicise the agri-food innovation agenda. Drawing on a socio-technical transition lens and a co-creation perspective, we identify three structural tensions that shape contemporary agri-food innovation: (i) regeneration versus extraction, (ii) participation versus capture, and (iii) data sovereignty versus data enclosure. We argue that the credibility of sustainability-oriented innovation depends less on the novelty of technologies than on how innovation is governed, whose knowledge counts, and how risks, costs, and value are distributed across actors and territories. We also introduce the contributions in this special issue, showing how they illuminate the politics of trust, legitimacy, participation, and public value in digital and institutional innovation across food consumption, governance, and producer engagement. We close with a research agenda centred on ethical experimentation, accountable AI, and mission-oriented governance that prioritizes planetary boundaries, distributive justice, and epistemic pluralism in agri-food transformation.

PurposeThe study investigates how entrepreneurs' experiences of hope and fear independently influence their entrepreneurial exit intention (EEI) amidst entrepreneurial adversity and examines how emotional ambivalence - the simultaneous experiencing of these intense contrasting emotions - affect EEI.Design/methodology/approachUtilizing linear regression analysis, the study empirically tests its hypotheses using survey data collected from 402 entrepreneurs in China.FindingsDrawing on the appraisal-tendency framework, the results uncover that hope and fear exert distinct effects on EEI, with hope decreasing and fear increasing EEI. Moreover, greater emotional ambivalence experienced during entrepreneurial adversity is associated with a heightened intention to exit entrepreneurship.Originality/valuePrior research on EEI has primarily focused on single emotions in isolation, overlooking the complexity of experiencing multiple conflicting emotions simultaneously, known as emotional ambivalence. This study addresses this gap by exploring emotional ambivalence, particularly the coexistence of hope and fear, which are closely linked to entrepreneurial uncertainty. By doing so, it offers valuable insights into emotional challenges entrepreneurs face when navigating adversity.

Although formal education is widely viewed as an important determinant of individual welfare, evidence on its relationship with entrepreneurs’ subjective well-being (SWB) remains mixed. Drawing on institutional theory, we examine how two types of country-level institutional supports- entrepreneurial education support and entrepreneurial funding support-influence the well-being returns associated with education-based human capital. Using multilevel data from 4,412 entrepreneurs across 20 countries and integrating the latest wave of the World Values Survey with institutional indicators from the Global Entrepreneurship Monitor and the World Bank, we find that these institutional supports moderate the formal education – SWB relationship in opposite ways. Specifically, entrepreneurial funding support makes formal education—SWB more positive, whereas entrepreneurial education support makes it more negative, such that highly educated entrepreneurs report lower SWB in countries with higher entrepreneurial education support. These findings underscore the functional heterogeneity of institutional support and clarify why prior evidence on the formal educationona – well-being nexus among entrepreneurs has been inconsistent.

This paper explores the optimal cross-border data transfer strategy for platform multinational companies (PMNCs). Using differential game theory, we model and compare three strategies: (i) a Nash non-cooperative game strategy, in which the government sets regulatory rules and PMNCs independently carry out cross-border data transfer; (ii) a Stackelberg leader-follower game strategy, in which the government plays a dominant role and uses subsidies to incentivize PMNCs to conduct cross-border data transfer; (iii) a cooperative game strategy, in which the government and PMNCs form a coalition to negotiate cross-border data transfer. The models incorporate the negative externalities of data elements, the value of cross-border data transfer, and the time factor. We study the optimal cross-border data transfer strategy and analyze the impact of key parameters on government and PMNCs decisions. The study reveals that: (1) The cooperative strategy between the government and PMNCs is the optimal option for cross-border data transfer management. (2) Government subsidies significantly promote cross-border data transfer for PMNCs and increase government revenue in turn. (3) Negative externalities generally hinder cross-border data transfer, while higher cross-border data transfer value stimulates this activity. The revenue distribution coefficient affects outcomes vary across different strategies. This study provides a theoretical foundation and practical guidance for PMNCs to determine their cross-border data transfer decisions and for governments to design effective regulatory and subsidy policies.

PurposeThe purpose of this study is to investigate the role of knowledge stock in the national-level emergence of tech scaleups from a configurational perspective. While prior research has recognized the importance of knowledge stock in the emergence of high-growth firms and examined elements of entrepreneurial ecosystems (EEs), its predominant reliance on econometric models has yielded a fragmented understanding, leaving unclear how knowledge stock combines with EE resource endowments to support such emergence.Design/methodology/approachDrawing on the knowledge spillover theory of entrepreneurship and ecosystem thinking, this study integrates aggregate data from the Global Entrepreneurship Monitor and Startup Genome. The authors construct a cross-national data set covering 31 countries and apply fuzzy set qualitative comparative analysis.FindingsThe results of this study reveal that knowledge stock plays a systemically important and consistently prominent role across high-emergence pathways; its influence is contingent upon the broader resource endowments of EE; and compared with resource endowments of EE, knowledge stock exhibits lower causal asymmetry, suggesting a relatively more stable role.Originality/valueThis study advances the literature on the geography of high-growth entrepreneurship by moving beyond single-factor econometric approaches to examine how knowledge stock interacts with other EE resources in national contexts. Using fuzzy set qualitative comparative analysis, the authors identify multiple resource configurations that enable tech scaleup emergence, reveal the context-dependent nature of knowledge stock's impact and demonstrate its lower causal asymmetry relative to other resources. These insights refine theoretical understanding of how different resource combinations jointly foster high-growth entrepreneurship.

We conducted a systematic review of the emerging literature on entrepreneurial failure stigma (EFS), synthesizing research published between 1999 and June 2025 across 48 articles. Our review refines the conceptualization of EFS and integrates previously intertwined but thematically divergent studies into a coherent theoretical framework. We define EFS as a set of negative beliefs about individuals who are directly or indirectly associated with an entrepreneurial failure event. These negative beliefs stem from two primary sources: external societal judgment and/or internalized self-perceptions by the individuals themselves. Based on this definition, we develop an integrative framework that systematically categorizes antecedents, consequences, and coping strategies associated with EFS. By consolidating diverse streams of research, we highlight critical knowledge gaps and identify promising directions for future inquiry. This review advances theoretical understanding of EFS and offers practical implications for entrepreneurs navigating the challenges of EFS.

Although prior research acknowledges that small and medium-sized enterprises (SMEs) can turn failures into growth opportunities, the mechanisms through which failure analysis contributes to such growth remain underexplored. Grounded in organizational learning and dynamic capabilities theory, this study explores how failure analysis facilitates SME growth through the mediating role of dynamic capabilities and the moderating role of environmental dynamism. Drawing on survey data from 207 managers of China SMEs, the study employs linear regression and bootstrapping techniques to empirically test the proposed hypotheses. The results reveal that failure analysis significantly promotes SME growth, with dynamic capabilities—specifically, sensing, seizing, and reconfiguring—serving as key mediators. Furthermore, environmental dynamism positively moderates both the relationship between failure analysis and dynamic capabilities, and the indirect effect of failure analysis on growth via dynamic capabilities. Unlike previous research that focuses primarily on innovation or resilience, this study uniquely highlights the role of failure analysis in cultivating dynamic capabilities to drive SME growth.

PurposeThis study aims to examine the effects of co-founders' responses to failure on entrepreneurs' recovery from failure.Design/methodology/approachThis study uses time-lagged survey data collected from 203 matched Chinese entrepreneur-co-founder dyads and interviews with four pairs of entrepreneurs and co-founders.FindingsThe results indicate that co-founders' tolerance of failure has a stronger positive effect on entrepreneurs' recovery from failure than co-founders' reflections on failure. The interaction between co-founders' tolerance of failure and reflection on failure positively influences entrepreneurs' recovery from failure. Two emotion regulation strategies, cognitive reappraisal and expressive suppression, moderate the relationship between co-founders' responses to failure and entrepreneurs' recovery from failure. The implications of the findings for research and practice are discussed.Originality/valueBy linking co-founders' responses to failure with entrepreneurs' recovery from failure, this study enriches the antecedents of entrepreneurs' recovery from failure from the perspective of co-founders and contributes to the evolving emotion regulation literature and application of social cognitive theory in failure recovery.

New product development (NPD) project failure, a common occurrence for high-tech firms, serves as an important source of learning and breeds future success, thereby prompting scholars to investigate the mechanisms of learning from NPD project failure. Studies has revealed the cognitive and emotional mechanisms underlying learning, and has investigated the role of post-failure factors in affecting these mechanisms; However, the influence of pre-failure factors is under-developed. Drawing upon the risk-as-feelings perspective, our study fills this gap by investigating one crucial pre-failure factor - risk anticipation. Our study categorizes risk anticipation into the likelihood anticipation and the magnitude anticipation, and hypothesizes that these two dimensions of risk anticipation affect learning from project failure distinctively. Based on two-wave data collected from 232 project managers in high-tech firms, we find likelihood anticipation positively affects learning from project failure primarily through a cognitive mechanism, as indicated by a positive mediation effect of decision-making comprehensiveness. In contrast, magnitude anticipation negatively affects learning from project failure primarily through an emotional mechanism, as indicated by a negative mediation effect of decision-making comprehensiveness. We shed light on the risk management practices for NPD project in high-tech firms, and offers insights into the post-project learning that breeds future NPD project success.

Purpose The entrepreneurial process often cannot be explained by a single entrepreneurial theory. Instead, it is more likely the result of the interaction between various entrepreneurial behavior patterns and different environmental conditions. However, existing research has frequently overlooked the complexity inherent in the entrepreneurial phenomenon. Building on a configurational perspective, this study aims to examine how new ventures can use different behavioral configurations to achieve high performance amid various uncertain environments. Design/methodology/approach Based on the survey data from 143 new start-ups in China’s software industry, this study uses fuzzy-set qualitative comparative analysis (fsQCA). Findings This study jointly considers multiple entrepreneurial behaviors − causation, effectuation and entrepreneurial bricolage and different types of environmental uncertainty − state uncertainty, effect uncertainty and response uncertainty. The findings reveal three behavioral configurations for high/nonhigh new venture performance. Originality/value This study expands previous insights into the relationship between entrepreneurial behaviors and new venture performance from the perspective of configurational theory. Moreover, it offers new insights into the types of uncertainty, further refining our understanding of the uncertainties inherent in entrepreneurial activities.