Abstract Large language models (LLMs) show increasing potential for complex analytical tasks; however, their reliability in high-stakes domains such as construction contract interpretation remains insufficiently understood. This study presents a controlled, human-in-the-loop evaluation of how prompt-engineering strategies influence the quality and perceived reliability of LLM-generated risk analyses for New Engineering Contract (NEC4)–style clauses. Generative pretrained transformer GPT-4.1 was applied in a deterministic configuration to identify contractor risks across six synthetic NEC4-style Z-clauses using five prompt strategies: baseline, direct, role based, chain of thought, and guided prompting. Thirty outputs were evaluated under a blinded protocol against expert-derived reference risk lists by a mixed-expertise cohort, supporting exploratory comparison. Assessment combined descriptive quantitative analysis of Likert-scale ratings of accuracy, completeness, clarity, legal salience, and hallucination risk with qualitative thematic analysis. Results indicate that guided, clause-anchored prompting produced more structured, legally grounded, and traceable outputs, whereas less constrained strategies exhibited narrative drift and weaker clause linkage. These findings suggest that structured prompting may support more consistent clause-level interpretation under controlled conditions, although conclusions remain bounded by the exploratory design, limited sample, and synthetic dataset. The study contributes to understanding how prompt design shapes interpretability and verifiability in contract analysis. The clause-guided human-in-the-loop (CG-HITL) framework is presented as a conceptual approach to support structured use of LLMs in preaward contract risk assessment.
Ultra-precision machining of silicon underpins semiconductor, optical and MEMS manufacturing, yet atomistic predictions remain highly sensitive to the interatomic potential employed. Here, seven popular interatomic potentials of silicon were benchmarked for thermal, cohesive, stacking-fault and tensile properties, with five carried through detailed dislocation and nanometric-cutting analyses and compared against TEM characterisation of ground silicon. The predicted responses vary markedly: apparent one-phase thermal-instability temperatures span 2309–3427 K, the unstable stacking-fault energies associated with the first {111} glide-set Shockley-partial displacement range from 1.918 to 2.948 J m−2, and Young's moduli vary from 65.06 to 99.93 GPa. Fitted dislocation-core energies, i.e., Ec(110) differ by almost a factor of six, from 0.1109 to 0.6398 eV Å−1. These differences translate into distinct stress-accommodation pathways during cutting. ABOP (Erhart and Albe) and classical Tersoff revealed localised deformation within narrow transformed/disordered subsurface regions, Stillinger–Weber activates transient dislocation-mediated plasticity, Screened ABOP promotes extensive disorder-mediated accommodation, and Kumagai exhibits an intermediate response. TEM observations of ground silicon revealed coexisting amorphisation, {111} stacking faults and partial-dislocation activity, supporting a mixed-mechanism picture of machining-induced damage. The results show that predictive atomistic simulations of silicon machining require mechanism-specific potential selection rather than validation against a single bulk property.
This scoping study assesses practical aspects of electrokinetic (EK) biocementation of clay soil underneath a railway embankment ahead of upscaled testing to include a reduced-scale field pilot as an intermediate step towards subsequent pilot embankment treatment. It considers suitable field setups and performs Life Cycle Analysis (LCA) of biocementation by biostimulation of carbonic anhydrase (CA)-producing bacteria compared to hydrated lime slurry, if both treatments were implemented electrokinetically. LCA analysis was conducted using SimaPro software (version 9.6.0.1) with Ecoinvent database and bench-scale laboratory testing data. Electroosmotic flow modelling was performed to instruct on suitable setups and for estimates of power consumption towards the field application of 30 m of railway embankment and foundation soil. LCA indicated a considerable reduction in global warming if CA biocementation is used (0.00823 kg CO2 eq for biocement vs. 0.022136 kg CO2 eq for lime), and resource usage (7.06 & times; 10-5 kg Cu eq compared to 8.47 & times; 10-5 kg Cu eq for lime). Biocementation was more water-consuming compared to lime, as it involved multiple chemical solutions. Terrestrial acidification, aquatic eutrophication, and ecotoxicity were slightly higher for biocement, possibly due to system boundaries and processes assumed for material production. Further sustainability improvements would be possible if waste materials (e.g., captured industrial CO2) could be used. Field trials will be essential for validation, system optimisation, and advanced model calibration.
Innovation is transforming the global business landscape, with the banking sector playing a pivotal role in this evolution. This study explores the dynamic interplay between financial technology (Fintech), the digital economy, and economic governance and their collective impact on bank performance across 33 Asian countries from 2011 to 2023. Employing Generalised Structural Equation Modelling (GSEM) and leveraging extensive datasets from BankFocus (BvD), GFDD, WDI, and IMF (FDI, FAS), the analysis reveals that Fintech adoption significantly enhances bank performance, with the digital economy serving as a crucial mediating factor. Moreover, economic governance emerges as a vital moderating influence, shaping the effectiveness of Fintech in driving banking outcomes. These findings underscore the importance of robust digital infrastructures and effective governance frameworks in harnessing Fintech's potential, particularly in emerging markets. These findings also contribute to the literature on organisational excellence by illustrating how digital transformation supports continuous performance improvement, a key tenet of quality management in financial institutions.
PurposeThis study examines how artificial intelligence-augmented decision-making (AI-ADM) capability influences triple bottom line (TBL) sustainability performance in small and medium-sized enterprises (SMEs), investigating technological, human, and organisational antecedents, the moderating role of environmental volatility, and heterogeneous effects across firm sizes.Design/methodology/approachGrounded in dynamic capabilities theory, organisational learning theory, and the resource-based view, we develop a comprehensive theoretical framework tested using a hybrid approach integrating Partial Least Squares Structural Equation Modelling (PLS-SEM) with Artificial Neural Networks (ANN). Data were collected from 304 SME managers and IT professionals. Multi-group analysis (MGA) examines firm size heterogeneity.FindingsPLS-SEM results reveal that all seven antecedents significantly influence AI-ADM capability, with technology infrastructure (beta = 0.342) and managerial skills (beta = 0.324) demonstrating the strongest effects. AI-ADM capability significantly enhances environmental (beta = 0.467), economic (beta = 0.524), and social performance (beta = 0.398). Environmental volatility positively moderates these relationships. ANN analysis confirms technology infrastructure and managerial skills as the most critical predictors. MGA reveals that medium-sized enterprises derive significantly greater sustainability benefits from AI-ADM than small enterprises.Research limitations/implicationsThis study is based on cross-sectional data from SMEs in BRICS economies, which limits causal inference and may not fully capture the dynamic evolution of AI-enabled decision-making capabilities. Although the hybrid PLS-SEM, PLSpredict, ANN and fsQCA approach enhances analytical robustness, the findings may not generalise to large firms or non-emerging markets. Self-reported measures may introduce perceptual bias, and future studies could incorporate longitudinal data, objective performance indicators and multi-respondent designs. Expanding the model to include generative AI, supply-chain dependencies or digital ecosystem interactions would further strengthen theoretical development and broaden the study's empirical relevance.Practical implicationsThis study provides SME managers with a clear roadmap for leveraging AI-enabled decision-making capabilities to improve sustainability performance in turbulent environments. By demonstrating how digital orientation, data governance and AI adoption jointly strengthen decision quality, the findings offer managers practical guidance on prioritising investments in data infrastructure, organisational learning and digital culture. The hybrid PLS-SEM, PLSpredict, ANN and fsQCA results reveal multiple capability pathways that SMEs can follow depending on their resource constraints. Policymakers can also use these insights to design targeted capacity-building programmes that accelerate digital transformation across emerging-economy SMEs.Social implicationsThe study shows that AI-enabled decision-making can enhance the social and environmental performance of SMEs, particularly in resource-constrained emerging economies. By improving transparency, data quality and evidence-driven choices, AI capabilities support fairer labour practices, community engagement and responsible resource use. The results underscore the potential of digital technologies to reduce inequality between digitally mature and digitally marginalised firms, enabling more inclusive participation in global value chains. Strengthening SME digital capabilities, therefore has broader societal benefits, including job preservation, skills development, improved environmental stewardship and greater resilience of local economies facing volatility and uncertainty.Originality/valueThis study extends dynamic capabilities theory to the AI-sustainability nexus, identifies resource configurations enabling AI-ADM in SMEs, demonstrates contingent effects of environmental volatility, and reveals firm size heterogeneity in AI-driven sustainability outcomes. The hybrid PLS-SEM-ANN methodology advances analytical rigour by capturing both linear and non-linear relationships.
