Working with robots: how employee–robot relationships shape employee retention

Developments in artificial intelligence (AI) have accelerated the integration of service robots into the service industry, including tourism and hospitality (Liu et al., 2025a; Gong, 2025). Robots can perform various service and hospitality tasks, in areas such as airport security, restaurant chefs, butlers, receptionists and concierges (Mukherjee et al., 2023). Growth is anticipated in the application of service robotics, with the hospitality industry being an early adopter of AI-enabled technologies (Gong and Huang, 2025). Organisations deploy robots to reduce operational costs, enhance efficiency and improve service quality (Guan et al., 2025). However, successful integration depends on their technical capabilities and upon employee adaption to the technology (Gong, 2025).

The hospitality industry faces challenges in relation to employee retention, recruitment and turnover due to the complex psychological, cultural and physical factors of the work environment (Zhao et al., 2025). It is also recognised that a skilled and stable workforce can underpin the long-term competitiveness and growth in hospitality (Goh and Lee, 2018). Robots in hospitality may exacerbate retention issues due to robot-phobia, associations with job insecurity and stress (Chen and Cai, 2024). Few studies have examined employee retention in association with the introduction of service robots or their influence on employees’ organisational attitudes. In the service context robots are also not only perceived as tools. Unlike manufacturing industries, where research has focused on safety, proximity and motion (Lasota and Shah, 2015), service robots can be seen as colleagues (Nyholm, 2024) and can influence workplace relationships.

Employee–robot interaction research has focused on technological acceptance and task performance (Marangunić and Granić, 2014). Research focus is now turning to relational aspects such as how employees frequently assign social roles to robots and build social relationships during human–robot interaction (Li et al., 2025). It has been indicated that individuals can develop real connections and feel less isolated through interactions with technology that conveys social and human-like qualities (Liu et al., 2025a). The literature, however, remains characteristic-centric, treating employee reactions as relatively uniform, such as robot usage and impacts on turnover intention (Zhang et al., 2023b). How employees perceive robots and through what mechanisms distinct employee–robot relationship types influence organisational outcomes remains theoretically underexplored.

Human-human organisational research suggests workplace relationships are crucial for organisational success (Dindia, 2003). As employees spend considerable time interacting with colleagues (Caillier, 2017) workplace relations are known to impact upon employee productivity and organisational performance (Janiukštis et al., 2024). These factors underscore the importance of studying employee–robot interactions from a relational perspective in addition to technological adoption and adaptation research. Within the employee–robot interaction literature, an organisational behaviour approach to employee–robot relationship and employee retention is limited. Therefore, this study attempts to answer two research questions:

Cultural contextual differences can influence the nature and depth of human–robot relationships (Bourne, 2025) and may shape how employees perceive and respond to robots (Shin et al., 2025). Human perceptions of robots can be formed by robot characteristics, but are also influenced by employee social norms, values and beliefs (Kou and Zhang, 2024). The study focuses on robot use and associated employee organisational attitudes but acknowledges that the cultural context may also influence perceptions.

The study conceptualises robots as organisationally embedded relational actors within the workforce. The integration of these human-machine hybrids can reshape patterns of reciprocity. From a signalling perspective, the deployment of robots may represent a structural decision that communicates organisational priorities to employees. Employee responses to robot integration, however, can be nuanced. Social information processing theory (SIP) argues that employees interpret relationships with robots as salient environmental cues that inform perceptions of role stability and organisational intention. Social exchange theory also suggests these interpretations shape employee-organisation exchange-based evaluations, which can influence job retention intention. These approaches offer a complementary framework to understand human–robot interaction from an organisational behaviour and relational perspective. Using a scenario-based experiment, this study seeks to explain the influence of employee–robot relationships on employee retention.

Service robots offer operational efficiency in tourism and hospitality through their machine learning capabilities. Robots can reduce operational costs and improve service delivery (Skubis, 2024). The integration of AI has enabled robots to be more flexible in their responses and adaptable to their environments (Ivanov and Webster, 2019). It has also increased robot capacity to perform social activities and make autonomous decisions (Lu et al., 2019). Research suggests that robots can be considered colleagues when sharing similar work content or domain of activity, have the same affiliation, or are a part of the same organisation (Nyholm, 2024). This makes it more common for robots to referred to as co-workers, team members or colleagues (Nyholm and Smids, 2019). This perception of human–robot interaction suggests robots are both tools and relational actors.

Robots may also function as organisational signals. Signalling theory posits that when one party has less information than another, the disadvantaged party uses signals to infer missing information and reduce asymmetry (Spence, 1973). These signals can be intentionally, or unintentionally, communicated by the signaller to the receiver, and characterise unobservable qualities about a product, service or organisation (Zhu and Zhang, 2023). Organisations reduce information asymmetry by providing credible and verifiable signals about attributes that receivers cannot directly observe (Lin et al., 2025). Beane (2020) suggests that robots can serve as visible signals in human–robot interaction. They can shape employee perceptions of organisational innovation and quality beyond their functional utility. Pan et al. (2024) also suggests that employees who observe organisational investment in robot development and witness customer satisfaction may engage in withdrawal behaviours at work. Employee collaboration or complaint intention may also be influenced by the introduction focus, employee serving or business (Huang et al., 2025).

An attribute-centric perspective has been extended in this study to conceptualise robots as relational agents to employees and present a proxy for organisational intent. Through this perspective employees develop relationships with robots. Employees may also infer organisational culture, values and managerial priorities from the way robots are designed, deployed and interact during work activities.

Human–robot interaction research suggests there are four functional relationships when working with robots Habib et al. (2026). In a collaborative relationship, employees perceive robots as partners who share goals and are interdependent (Le et al., 2022). A competitive relationship occurs when employees perceive robots as a replacement and a threat to resources and opportunities. This relationship results in employees adopting a zero-sum mindset (Jin et al., 2025). A supplementary relationship may develop where there are similarities between employee and robot characteristics, such as values and goals. Within a complementary relationship, one party provides what the other wants and operates as a mechanism for need satisfaction (Ryu, 2014).

Organisational research applies social information processing theory (SIP) to explain the factors that influence employee behaviour (Lu et al., 2018). The theory posits that social environments significantly influence individual behaviour and reactions (Salancik and Pfeffer, 1978). It suggests an individual’s social environment conveys information that shapes their perceptions, attitudes and behaviours. By processing contextual information, individuals may gain a better understanding of their context, which in turn shapes their responses (Yang et al., 2025).

Social information, such as coworkers’ observations, behavioural experience or comments, may provide cues for employees to construct and interpret work events (Ling et al., 2025). This study posits that employee perceptions of organisational and managerial intentions are shaped by a variety of social cues derived from their relationships with robots-coworkers. A collaborative robot may convey managerial intention to improve the workplace for employees, while a competitive relationship may signal efficiency and profit at the expense of human employees.

In human–robot interaction, humans treat robots as machine-human hybrids who are alive enough to be anthropomorphised but also machine enough to be treated as not-quite human (Bankins and Formosa, 2024). Kahn et al. (2004) argue that apparent reciprocity in human–robot interaction remains one-sided, with humans retaining control and freedom from moral obligation, as robots do not anticipate equivalent reciprocation when providing support (Yu et al., 2026). Similarly, although focused on customers, Belanche et al. (2020) found that responsibility for service failures is more often attributed to the organisation than to frontline robots.

Viewing robots as a human-machine hybrid, relational actors and organisational signals suggest that the reciprocity process may operate asymmetrically compared to traditional human-human interaction. Employee social exchange evaluations are therefore likely to be directed towards the organisation that deploys the technology. While reciprocal dynamics with robots may emerge, this study focuses specifically on the employee-organisation social exchange process.

Social exchange theory (SET) is commonly applied to explain employee behaviours and attitudes in the workplace (Nguyen et al., 2023). It reflects human–human relations, such as employee-leader and employee-coworker exchange (Zhu et al., 2023). It may also reflect the employee–organisation relationship. Studies have applied SET to examine how the organisation treats the individual, how the individual responds and how this culminates in a mutually beneficial exchange (Le et al., 2024). SET posits that social interactions are shaped by exchanges that maximise benefits and minimise costs. Employee perceptions of organisational support and value foster reciprocal positive attitudes and behaviours such as loyalty, engagement and enhanced performance (Al-Romeedy et al., 2024). Using SET, this study views the employee–robot relationship as an extension of the employee-organisation reciprocal relationship.

Employee retention is a critical managerial concern and understanding the catalyst for employee loss is important for organisational success (Utete, 2023). Research has indicated that peer relations affect employee retention. Positive coworker relationships can increase employee motivation and intention to stay (Basford and Offermann, 2012). Poor or conflicted relationships may cause rivalry that negatively impacts on work quality and may result in the loss of talented employees (Gelencsér et al., 2023). From an employee–robot relationship perspective, employee relationship type may influence retention through (SIP).

A positive work environment and perceived organisational support may also influence employee retention (Arasanmi and Krishna, 2019). Rather than producing a uniform effect, robot use may influence employee retention through competing pathways. Zhang et al. (2023b) found that robot use reduced hotel employees’ turnover intention through work autonomy but increased it through job insecurity. Where employees have no voice in the adoption of robots, a defensive position may be triggered. These defensive positions may induce feelings such as the fear of being replaced by robots (Sun and Deng, 2024). These studies align with the perspective of signalling between organisations to employees.

Drawing on social exchange theory, an employee–robot relationship may influence employee retention as reciprocal behaviour from a SIP and signalling perspective. Therefore, the following hypothesis was proposed:

According to information processing theory, employee attitudes are formed by social interaction with others in the workplace (Salancik and Pfeffer, 1978). Positive perceptions of work environments and cooperative relationships with coworkers and managers demonstrate high levels of job satisfaction (Westover and Taylor, 2010). Social exchange theory has also posited that workplace relations have a strong, direct link to job satisfaction (Abugre, 2017).

The presence of a robot in a work environment can affect employee satisfaction, depending on the nature of the work (Findlay et al., 2016). Employees’ perceptions of robots’ role may also matter, particularly whether robots are seen as supporting employees or reshaping the work environment (Nonaka et al., 2020). Satisfaction may also depend on whether, and to what extent, employees interact with robots (Turja et al., 2022) and whether the human or the robot leads (Pasparakis et al., 2023).

Positive and negative perceptions of robot coworker relationships and robot deployment intention influences job satisfaction. Job satisfaction is an important factor in employee turnover (Pasparakis et al., 2023). While employee–coworker relationships influence job satisfaction, they also influence turnover intention (Kim and Kim, 2017). A decline in job satisfaction increases the intention to quit, and vice versa (Maier et al., 2012). Hence, the following hypothesis was proposed:

Discussion of employee engagement has been founded on social exchange theory (Pariyanti et al., 2024), as reciprocal attitudes and behaviours towards coworkers or an organisation. Research shows that the quality of workplace relations strongly influences engagement. Poor relations can drive negative engagement (Kisamore et al., 2010) and energising relationships can replenish employee psychological resources, leading to more motivated, committed and productive employees (Owens et al., 2016). Perceived organisational support has also been positively associated with engagement (Harris et al., 2025). As employees connect with organisational values, they tend to be more engaged and devoted (Chowdhury et al., 2024), particularly when they sense psychological empowerment (Tan et al., 2025). Employees may also perceive organisational values through interactions with robots. For example, a competing robot that restricts employee psychological empowerment may affect employee social exchange towards the organisation and robots.

Hospitality industry employee retention studies show that employee engagement influences employee intent to remain with an organisation (Karatepe et al., 2020). Robots have also been perceived as resources that can assist or detract from workplace outcomes (Liu et al., 2025b). This may be evident when an organisation invests in resources that foster a sense of meaningfulness and safety, which enhances engagement and retention (Yulita et al., 2020). Based on this premise, the following hypotheses were proposed:

Hospitality work is often performance-driven, requiring employees to meet productivity and service targets. This can create uncertainty and a sense of job insecurity, especially among entry-level employees (Ghani et al., 2022). The implementation of robots can increase the fear of employee replacement, especially for low-skilled workers (Chao and Kozlowski, 1986). Job insecurity can also be influenced by conflict between co-workers (Vásquez et al., 2018) or engagement in social comparison and competition with robots. Tojib et al. (2023) has highlighted that the robot design affects employee perceptions of a robot’s ability to replace their jobs, which leads to a sense of insecurity. Wang et al. (2023) indicated that humanlike features in robots may trigger employees to compare their skills and performance against robots. This may increase the fear of replacement. These studies support an organisational signalling perspective and the meaning of robot presence to employee morale.

While robots bring many benefits to the workforce, employee–robot interaction research indicates that robots may be associated with lower job satisfaction and serve as a more refined form of management control (Kowalski, 2025). Robot-induced job insecurity may therefore influence organisational outcomes such as employee turnover, job satisfaction and engagement (Moy et al., 2022). Subsequently, the following hypotheses were proposed:

Drawing on Novak and Hoffman (2018) and Habib et al. (2026), this study used an experimental design. This design aimed to examine participants’ reactions to service robots in a restaurant setting and was controlled for age, gender and education. A total of 347 valid responses from service industry employees were collected via Prolific. Participants were compensated at £6 per hour. On average, the survey completion took 11 min. This study adhered to the ethical research guidelines for studies involving human participants and was approved by the university’s Ethics Review Board. Participants were informed of the study purpose, the voluntary nature of participation, their right to withdraw at any stage without consequence and the measures used to protect their anonymity and confidentiality. Informed consent was obtained prior to response submission. Data were securely stored and used only for academic research purposes, and the study involved no foreseeable risks beyond those normally associated with scenario-based experimental research.

Participants were randomly assigned to one of four scenarios and asked to read as if they were the employees in that scenario. Robot agency and communion were manipulated in the scenarios to establish the four relationship types (collaboration, competition, supplementary and complementary) and test employee outcomes. Participants were asked to answer manipulation-check questions and to report their perceptions of the four types of relationships and outcomes (Habib et al., 2026).

All variables were measured on seven-point Likert scales. The items were adopted from previous studies. The collaboration scale was adopted from Zhang et al. (2023a) and measured by five items. Competition was measured by three items from Gordils et al. (2023). Supplementary and complementary relationships were measured by five items each from Piasentin and Chapman (2006). Employee engagement was measured across three dimensions and seven items (two items in each of the physical and cognitive dimensions, and three items in the emotional engagement dimension), adopted from Koo et al. (2020). Job security threat was measured by two items adopted from Tojib et al. (2023). Intention to remain was measured using three items adapted from Koo et al. (2020). Job satisfaction was measured by three items adopted from Frye et al. (2019). The full list of measurement items is provided in Table A of the supplementary materials.

Manipulation checks confirmed that the agency and communion manipulations were effective. The subsequent analyses focus on how the four types of employee–robot relationships influence employee outcomes. This study used linear regression analysis to test H1. To examine mediation and serial mediation, Hayes’ SPSS PROCESS Macro Models 4 and 81 were used, with 5,000 bootstrap samples and 95% bias-corrected confidence intervals to test the proposed hypotheses (Muñoz et al., 2024). Results are presented in Tables 1 and 2. Age, gender and education were used as demographic control variables in all analyses (Lian et al., 2025). The inclusion of control variables did not alter the substantive pattern of results. To account for the coexistence of relationship types (De Cooman et al., 2015), all models controlled for the remaining employee–robot relationships, allowing the unique effect of each focal relationship to be estimated.

The results revealed that employee-robot collaborative, supplementary and complementary relationships had a significant positive effect on intention to remain. In contrast, competitive relationships had a significant negative effect, thereby supporting H1a–d. The perception of the robot as a competitor appears to be harmful for retention, whereas positive relationship perceptions increased retention.

The results indicate that job satisfaction plays a central mediating role in linking employee–robot relationships to employees’ intention to remain. The indirect effects of collaborative, supplementary and complementary relationships on intention to remain through job satisfaction were positive and significant. The corresponding direct effects became non-significant after job satisfaction was included. Thus, H2a, H2b and H2c were supported with full mediation, suggesting that positive employee–robot relationships enhance employee retention through increased job satisfaction. Job satisfaction also mediated the relationship between competitive relationships and intention to remain. The indirect effect was negative, and the direct effect remained significant, indicating partial mediation that supported H2d.

Employee engagement also mediated some, but not all, employee–robot relationship effects. Collaborative relationship had a positive significant indirect effect through engagement, with a non-significant direct effect, supporting H3a with full mediation. Supplementary relationship did not show a significant engagement-mediated effect, therefore, H3b was not supported. Complementary and competitive relationships showed significant indirect effects through employee engagement, with direct effects remaining significant. This indicates partial mediation, with a positive engagement-mediated effect for complementary relationship and negative effect for competitive relationship, supporting H3c and H3d.

The serial mediation results show that job security threat links certain employee–robot relationships to intention to remain through job satisfaction and employee engagement. Collaborative relationship had significant positive serial indirect effects through job security threat, followed by job satisfaction and employee engagement, supporting H4a and H5a. In contrast, supplementary and complementary relationships showed no significant serial indirect effects, therefore, H4b, H4c, H5b and H5c, were not supported. Competitive relationships had significant negative serial indirect effects through job security threat, job satisfaction and employee engagement, supporting H4d and H5d. Thus, collaboration appears to reduce threat and competition increases perceived threat.

Drawing on De Cooman et al’s. (2015) suggestion that where team members share higher perceptions of supplementary and complementary fit, they will outperform those with less higher perceptions. Post hoc analyses were conducted to investigate relationship types coexistence and the nature of their effect. Positive relationship breadth was calculated as the number of positive relationship types rated above the scale midpoint. Mean outcome scores were then compared across participants perceiving zero, one, two or three positive relationships. The results supported the coexistence and cumulative value of positive relationship perceptions. Participants varied in the number of positive relationship types they perceived, and higher positive relationship breadth was associated with higher mean scores for intention to remain, job satisfaction and employee engagement.

To examine whether competitive relationships moderated the effect of positive employee–robot relationship perceptions, moderation analyses were conducted with job security threat, job satisfaction, employee engagement and intention to remain as outcome variables. Hayes’ SPSS PROCESS Macro Model 1, with 5,000 bootstrap samples and 95% bias-corrected confidence intervals, controlling for age, gender and education. All models also controlled for the remaining employee–robot relationship types, allowing the unique effect of each focal relationship to be estimated. The results showed that positive and competitive relationship perceptions did not interact, suggesting an independent effect rather than interactive pattern. The post hoc analyses suggested that different relationship types perceptions can coexist. More positive relationships can enhance outcomes, while competitive relationship perceptions contributed negatively and independently, rather than moderating the effect. Employee outcomes by positive relationship breadth are presented in Figure 2.

This study examined how employee–robot relationships influence employee retention. The findings show that they do not influence employee retention in a uniform way, rather, they operate through distinct psychological pathways. This suggests that employees can respond to robots positively or negatively, but also according to the specific relational meaning attached to the robot. These findings complement research that has tended to approach employee–robot interaction outcomes as being relatively uniform (Zhang et al., 2023b).

The findings suggest four distinct relational pathways. A supplementary relationship operates mainly through similarity, fit and satisfaction, when the employees feel that the robot is compatible with them and the workplace. Under these circumstances employees evaluate their job more positively. A collaborative relationship operates through interdependence and reassurance. When the robot collaborates with employees, it increases employee satisfaction, engagement and reduces perceived threat. Complementary relationships operate through capability extension, when the robot provides job skills the employees do not have, thereby strengthening satisfaction and engagement. Competitive relationship operates through threat, where the robot is perceived as rival. Under these conditions, employees experience insecurity, lower satisfaction and engagement and weaker retention. These interpretations align with the different relationships’ characteristics (Le et al., 2022; Jin et al., 2025; Ryu, 2014).

Job satisfaction emerged as the dominant mechanism linking employee–robot relationships to employee retention. Positive relationships enhanced intention to remain through higher satisfaction, suggesting that robots were perceived as collaborative, similar or capability-enhancing. These perceptions made the job feel more manageable and supported. Competitive relationships showed the opposite pattern, reducing satisfaction and retention intention. With positive relationships signalling organisational support, the findings align with research showing the impact of organisational support on employee satisfaction (Sarıkaya and Kara, 2020).

Employee engagement offered a more selective pathway. While satisfaction explains whether employees evaluate their job more positively, engagement explains whether the robot relationship activates employees’ involvement in work. Engagement mediated collaborative and complementary relationships but not a supplementary relationship. Collaboration may activate engagement through joint work, and interdependence, while complementarity does so by extending employee capabilities. Supplementary relationships appear to create fit and comfort rather than active involvement. Competitive relationships reduce engagement as the robot is perceived as a rival rather than a resource, weakening employee motivation and retention.

The serial mediation results show that job security threat plays a central role in the negative effect of competitive relationship. Competitive relationship increased job security threat, which reduces job satisfaction and engagement, lowering intention to remain. This indicates that competition operates through a cascading psychological process, where employees who perceive the robot as rival feel more insecure. This insecurity weakens employee satisfaction and work involvement. A collaborative relationship shows the opposite pathway by reducing job security threat and supporting retention through satisfaction and engagement. This indicates that the interdependent nature of collaboration adds assurance, which reduces job insecurity. Supplementary and complementary relationships, however, did not significantly operate through job security threat. While research has highlighted job insecurity perceptions in working with robots (Tojib et al., 2023), findings in this study extend the discussion by adding the different relationship types and their impact on job insecurity.

Additional analyses suggested that employee–robot relationship perceptions can coexist, which aligns with research (De Cooman et al., 2015). Employees may interpret the same robot through multiple positive relational frames, and positive relationship breadth was associated with stronger retention, satisfaction and engagement, but not with lower job insecurity. This suggests that positive outcomes such as working with employees, fitting the workplace and extending employee capabilities, can support favourable outcomes without removing threat perceptions.

The moderation analysis further showed that competitive perceptions did not significantly moderate the effect of positive perceptions. Instead, the findings suggest an independent rather than interactive pattern. Positive perceptions contributed to favourable outcomes, while competitive perceptions independently contributed to less favourable outcomes and higher job insecurity. Thus, positive framing alone may not neutralise competitive concerns. Multiple relationship perceptions can coexist. Positive relationship breadth shows cumulative associations with favourable outcomes, while competitive perceptions remain an independent source of risk overall.

This study extended employee–robot interaction research by showing that robots are interpreted through relational meanings that influence employee retention. Research has focused on technology acceptance, usefulness or threat (Marangunić and Granić, 2014; Chao and Kozlowski, 1986). The findings of this study show that employees’ responses to robots are also shaped by the type of relationship they perceive with the robot. This shifts the attention from whether employees accept robots to how employees relationally position robots in the workplace.

The study contributes to organisational behaviour and retention research by identifying distinct psychological mechanisms linking employee–robot relationships to intention to remain. Job satisfaction emerged as dominant attitudinal pathway, employee engagement as a selective motivational pathway and job security threat as a threat-based pathway. This shows that employee–robot relationships influence retention through different internal mechanisms rather than through a single general response.

The findings of this study contribute to fit and relational theories by demonstrating that positive relationship types are not interchangeable. While all may lead to satisfaction, supplementary relationship reflects fit and similarity. Collaborative and complementary relationships enhance engagement through their nature of interdependence, and reassurance for collaborative and capability extension for the complementary. These distinctions refine an understanding of how positive relationships with non-human workplace actors operate.

Finally, this study contributes to theory by confirming that employee–robot relationship perceptions can coexist and operate independently. Employees may hold multiple relational meanings towards the same robot. Rather than treating relationship types as mutually exclusive categories, the findings suggest that employee–robot relationships are multifaceted.

This study provides valuable practical implications. Service robot implementation should be managed as a technological deployment (Go et al., 2020), but also as a relational process. Employees respond to what robots do, but also to what robots mean in relation to the employees’ own work. Managers and policymakers should focus on how employees interpret robot coworkers and the potential outcomes of these relationships.

Organisations can strengthen retention by signalling robot deployment to enhance job satisfaction. Positive relationships enhanced retention through satisfaction. Managers and human resource practitioners can frame robots as resources that improve the employee work experience rather than efficiency tools. Communication should emphasise how robots support work conditions, reduce burdens, align with service values and preserve the value of human roles.

Robot deployment can cultivate distinct and multiple positive relational cues. Collaborative cues can be developed through workflows that position robots as working with employees rather than replacing them. Supplementary cues can be strengthened by aligning robot behaviour, tone and service style with workplace norms and service culture. Complementary cues can be reinforced by clarifying how robots extend employee capabilities. The cumulative nature of positive employee–robot relationships on retention suggests that managers and designers can focus on embedding more than one positive relationship type to enhance retention.

Organisations need to manage competitive perceptions directly. Competitive relationship represents an independent retention risk and can undermine satisfaction, engagement and retention. Managers, human resource practitioners and technology providers should avoid messages that compare robots with employees or imply replacement. Instead, implementation should emphasise role clarity, reskilling opportunities, career pathways and the continued importance of human employee roles. The goal is to cultivate multiple distinct positive relationships while minimising competitive threats. Finally, human resource managers can monitor workplace relationships with robots and their influence on organisational attitudes.

Despite efforts to ensure rigor, this study has several limitations. Generalisability may be constrained by data collected from within Australia, the UK, Spain and the USA, indicating the need for broader cross-cultural validation. The use of a scenario based experimental design, while appropriate for an emerging field, highlights the importance of future triangulation and longitudinal research in human–robot interaction. Although frontline employees were included, only a minority worked in hospitality. Despite consistent results across subsamples, replication with larger, industry specific samples are warranted. While adapted measures demonstrated acceptable reliability and validity, they were not independently validated in an employee robot context, suggesting the need for further scale development.

Future research could extend this study by examining reciprocity towards robots as they become more autonomous. Studies could also identify additional employee–robot relationship types and examine boundary conditions such as employee role, technology experience, employment status, perceived replaceability and task overlap with robots. For example, competitive perceptions may be more harmful for employees who perceive fewer alternative job opportunities. Future studies could further distinguish similarity, fit and substitutability by examining when supplementary relationships signal compatibility with employees’ work and when they raise concerns about replacement risk. Finally, more refined methods could be developed to examine relationship profiles and test how combinations of positive and competitive meanings shape employee outcomes over time.

The authors acknowledge that this paper uses the same data set as our previous work. However, the present study addresses different research questions and examines different outcome variables.

This information is just for review process and not part of the online publication acknowledgement section.

The supplementary material for this article can be found online.