Retailers increasingly deploy generative artificial intelligence (GenAI) to handle complaints, yet evidence remains limited on when these agents restore purchasing rather than merely improve immediate evaluations. We develop a configuration model in which recovery agent identity (GenAI versus human) is separated from recovery agent empowerment, defined as authority to complete a remedy during the encounter. Across a qualitative study, two pretests, two randomised experiments, and two secondary-data studies in United Kingdom retailing, we examine perceived recovery justice, customer effort reduction, and trust restoration as mechanisms linking recovery configurations to subsequent behaviour. The experiments capture an incentivised purchase choice, whereas the field studies observe repeat purchase, spending, time to next purchase, and churn. GenAI was not inherently superior to a human agent. Its advantage emerged when it had visible and executable authority to issue refunds, replacements, delivery upgrades, or account credits. Empowered GenAI improved justice, reduced effort, and restored trust, and generated the strongest purchase outcomes. Failure severity weakened this advantage, while customer relationship strength amplified both the penalty for low-empowerment GenAI and the reward for empowered GenAI. The study contributes an authority-identity decoupling perspective: customers respond not only to who communicates, but also to whether the recovery architecture gives that agent the data access, decision rights, and operational capability required to act. The findings support selective empowerment for routine failures and accountable hybrid escalation for severe or data-constrained cases.
This study explains when eco-certification creates firm value in hospitality and why similar certification programmes generate different outcomes across institutional settings. Using a balanced panel of 50 listed hospitality firms from the United Kingdom, the United States, and Egypt (2015–2025; 550 firm-year observations), the study combines staggered difference-in-differences estimation with implementation-alignment analyses. Results show that eco-certification intensity increases both firm value and green innovation, while green innovation transmits most of the valuation effect. Additional analyses identify a certification-translation gap, whereby firms with extensive certification coverage but limited innovation receive significantly lower valuation benefits. Certification effects are nonlinear and delayed, becoming stronger only after certification reaches meaningful portfolio breadth. Regulatory stringency and environmental norms mainly stimulate implementation, whereas sustainable finance depth and governance quality facilitate and capitalise innovation. The study reconceptualises eco-certification as a translation-governance mechanism whose value depends on implementation alignment, scale, timing, and institutional support.
This study examines whether artificial intelligence enabled waste to energy adoption improves financial performance in tourism, travel, and hospitality firms and identifies the mechanisms and conditions shaping these effects. Using an empirics first mixed methods design, we combine interviews with 41 managers and a firm year panel of United Kingdom listed tourism, travel, and hospitality firms from 2015 to 2025. Difference in differences models with firm and year fixed effects show that adoption is associated with significantly higher financial performance, particularly at greater levels of adoption intensity. The findings indicate that energy cost savings and operational efficiency partially explain these gains. Performance benefits are also stronger in firms with higher baseline waste intensity and during periods of elevated energy prices. The study demonstrates that AI enabled waste to energy functions as an operational capability that can generate measurable economic value when integrated into organisational routines and governance systems.
Over the past two decades, research into Internet Gaming Disorder (IGD) has markedly increased due to worldwide spread of online videogames. The reasons and motivations for playing greatly contribute to its popularity. Escapism and avoidance coping strategies have been studied extensively and conceptualized as motives to play. A growing research base has demonstrated a strong association between these motives to play and negative gaming outcomes. Consequently, the aim of the present systematic review was to provide a comprehensive overview of the role of avoidance coping and escape motives in problematic online gaming. A systematic literature search was carried out using academic databases and a total of 26 empirical studies met the inclusion criteria. The results show that escapism and avoidance coping represent both a predictor of IGD and play a mediating role between many psychological factors (e.g., self-esteem, loneliness, self-concept, anxiety) and problematic online gaming. However, the review also highlights the paucity of longitudinal studies that hinder the determination of the causal direction of these associations. Despite this limitation, the evidence has important implication for developing more effective prevention programs and clinical interventions.
Robust deployment of large multimodal models (LMMs) in real-world scenarios requires access to external knowledge sources, given the complexity and dynamic nature of real-world information. Existing approaches such as retrieval-augmented generation (RAG) and prompt engineered search agents rely on rigid pipelines, often leading to inefficient or excessive search behaviors. We present MMSearch-R1, the first end-to-end reinforcement learning framework that enables LMMs to perform on-demand, multi-turn search in real-world Internet environments. Our framework integrates both image and text search tools, allowing the model to reason about when and how to invoke them guided by an outcome-based reward with a search penalty. To support training, We collect a multimodal search VQA dataset through a semi-automated pipeline that covers diverse visual and textual knowledge needs and curate a search-balanced subset with both search-required and search-free samples, which proves essential for shaping efficient and on-demand search behavior. Extensive experiments on knowledge-intensive and info-seeking VQA tasks show that our model not only outperforms traditional RAG-based baselines of the same model size, but also matches the performance of a larger RAG-based model while reducing search calls by over 30%. We further analyze key empirical findings to offer actionable insights for advancing research in multimodal search.
