Abstract

Robots have evolved from industrial machines designed to perform hazardous, repetitive tasks to social companions capable of interacting with and assisting humans in healthcare, education and daily life. Social robots interact with humans using verbal, non-verbal and emotional cues, offering support and companionship while requiring social intelligence and ethical safeguards for effective and trusted use. AI-enhanced robots in education improve learning, engagement and problem-solving skills, particularly in STEM subjects, while also supporting inclusive and personalised instruction. However, careful implementation is needed to address emotional attachment risks and avoid dehumanising the learning experience. AI-powered robotics in healthcare enhance diagnostic accuracy, surgical precision, personalised medicine and emotional support for patients. However, challenges such as algorithmic bias, privacy risks, economic impacts on medical staff and ethical concerns require careful oversight, robust safeguards and inclusive data particles to ensure equitable and safe healthcare delivery. AI-enhanced robots raise complex ethical and social challenges, including concerns about bias, privacy, consent, accountability and the potential for overreliances or social displacement in sensitive contexts like healthcare caregiving. AI-enhanced robots have the potential to provide companionship, emotional support and personalised interactions, strengthening human-robot bonds. At the same time, ethical considerations, privacy, safety and societal trust set important boundaries for their use.

1. Introduction

The definition of a robot is a programmable machine that is capable of carrying out a series of tasks autonomously (OED, n.d.). Robotics has historically been intertwined with industrial use, however the use of robots has expanded past factories and is now being implemented into various fields.

The first ever industrial robot, Unimate, was developed in the late 1950s and conceived from a design for a mechanical arm patented in 1954. It was designed entirely for industrial use, its purpose being to replace humans in hazardous, repetitive factory jobs. Unimate started the industrial robotics revolution, showing that robots were able to improve safety, productivity and precision; it was simply time efficient. By the 1970s, many car factories worldwide were using Unimation robots. The use of robots was predominantly prevalent in industrial areas until the 1960s and 1970s, when the first prosthetic arm was produced in 1968 (Cyberne1, 2015). This expanded the utility of robots and allowed mechanics to flourish in unexpected areas such as the medical field, where artificial intelligence and robotics were combined to help prescribe patient medication. The impact of this development can be seen today in ARI – a full-sized humanoid robot specialising in care for elderly patients (Gareth Willmer, 2025). Willmer conducted a study with 100 patients and found the robot elicited a postive response, demonstrating a drastic shift in the societal role of robots since their induction.

This shift is reflected in the growing field of social robotics. Social companion robots are designed specifically for human interaction, companionship and care (Prescott & Robillard, 2020). For example, Pepper is a humanoid social robot created by SoftBank Robotics. First unveiled in 2014, Pepper was designed to fully interact with humans on an emotional and social level (Sievers, 2018). It was recognised as one of the first commercially-available robots aimed at companionship and customer engagement rather than industrial work. Pepper had a variety of applications as it was utilised in sectors such as education, healthcare and even customer service. 

2. Social Robots

2.1 What is a Social Robot?

Social robots are autonomous or semi-autonomous systems that interact and communicate with humans using socially appropriate behaviours (Onyeulo & Gandhi, 2020). Their design often includes anthropomorphic or zoomorphic elements such as expressive eyes, facial features and body language to foster engagement and social connection (Onyeulo & Gandhi, 2020). These robots often combine speech, gaze gestures, facial expressions and even haptic cues to fit intuitive communication. A key element in their design is the use of expressive modalities that simulate empathy or emotional states, as it makes users feel more comfortable and connected during interactions (Bishop et al., 2019). For instance, robots with dynamic and positive facial and vocal expressions might be better aligned with human expectations of a natural conversation. 

In many cases, the effectiveness of a social robot depends not on its technical sophistication alone, but on how well it is perceived as socially competent. Farouk (2019) notes that some robots that prioritise non-verbal communication, such as gaze and posture, often  carry greater weight in shaping human impressions during short-term interactions. Emotional intelligence, defined as the ability to recognise, interpret and respond to human emotions, is another central characteristic, especially in healthcare, therapy or companionship applications where emotional resonance is critical for user trust and long-term engagement (Farouk, 2019).

2.2 Importance and User Acceptance

The importance of social robots touches concrete social needs in areas such as elderly care, education rehabilitation and mental health support. Due to the fact that social robots are designed to be proficient in social skills, they can mitigate loneliness in ageing populations, provide consistent assistance to individuals with cognitive impairments or support students in learning environments through adaptive, interactive teaching methods (Farouk, 2019). Their roles are particularly relevant in societies facing shortages in human caregiving resources, where they serve as supplementary agents. 

User acceptance, however, is a pivotal factor in their successful deployment. As highlighted by Onyeulo and Gandhi (2020), the design of facial animations, speech patterns and interactive gestures strongly influences whether people find these robots approachable, trustworthy and engaging. Robots that align with cultural expectations and communicate in a manner consistent with local norms are more likely to be accepted (Onyeulo & Gandhi, 2020). Furthermore, participatory design approaches – where potential users, including older adults or patients, are involved in shaping the robot’s features – have been shown to enhance usability and perceived usefulness (Farouk, 2019). This user-centred approach ensures that the robots’ social competencies are not only technically functional but also contextually relevant.

2.3 Human-Robot Interaction and Multimodal Communication

Human-robot interaction (HRI) is the cornerstone of social robotics, encompassing the ways in which humans and robots exchange information, coordinate actions and establish social rapport. Effective HRI relies on multimodal communication strategies that integrate verbal language with non-verbal cues such as gaze, body language and facial expression (Onyeulo & Gandhi, 2020). Non-verbal elements often carry affective meaning and serve to enhance or even substitute for verbal communication, especially when the robot is designed for populations with cognitive or language limitations (Farouk, 2019). 

Sundar (2020) further expands this by introducing the concept of machine agency, where robots are not passive conduits of interaction but active agents capable of shaping user experience. Through adaptive behaviours, personalised responses and interactive affordances, social robots can guide users’ emotional states, decision-making processes and engagement levels. This capacity to influence interaction outcomes is central to the evolution of HRI, as it reframes robots from being mere tools to becoming collaborative partners in various social contexts.

2.4 Social Intelligence in Robots

Ghandi and Onyeulo (2020) highlight how social intelligence in robots is not solely dependent on human-like appearance, but also on their capacity to respond and interpret both verbal and non-verbal cues accurately. Through integrating social and conversational cues, robots may learn to build emotional bonds with humans (Gandhi & Onyeulo, 2020). These elements contribute to the robots’ abilities to convey intentions and emotions, creating more natural and relatable interactions. Additionally, this emphasises the significance of emotional intelligence in robots and their design, further aiming to create robots who can understand and respond to human emotions and social contexts. 

3. Impacts of AI-Enhanced Robots on Society

3.1 AI-Enhanced Robots in Education

Combining artificial intelligence with robotics has transformed traditional industries, including the education sector. AI-enhanced robots can support personalised learning, foster engagement and address gaps in one-to-one support.

The implementation of robots in education has proved to have a positive impact on student grades and performance, especially in STEM subjects. A detailed analysis across 21 separate studies revealed a strong improvement in learning performance and attitudes when AI-enhanced robots were used in the classroom (Ouyang & Xu, 2024). The study also showed that robots were particularly effective in technological fields, less so in more traditional areas like mathematics. The best results were presented when the robots actively supported learning, when lessons were structured via a project or game‑based approach and when interventions lasted over a month (Ouyang & Xu, 2024).

Many studies have documented that hands-on robot usage greatly boosts problem-solving skills, teamwork and conceptual understanding (Kaur et al., 2025). Educational robots also broaden computational thinking, creativity and positive learning habits, especially notable in the earlier years of education such as primary and secondary schooling (Selcuk et al., 2024.). Students also exhibited increased motivation and interaction when robots were used in their lessons or lectures (Selcuk et al., 2024).

Robots have also demonstrated the potential to engage students regardless of their likes and dislikes of a subject, increasing accessibility. In a study using the “Thymio” robot and Scratch to teach geometry to 15 year olds, students found robotics-based exercises more interesting and useful compared to traditional ones, regardless of whether they initially liked mathematics, highlighting strong potential for social robotics to broaden participation in STEM subjects (Brender, 2021). Moreover, robots support inclusive education: robots such as NAO and Kaspar provide open and equal interactions for students who are on the autism spectrum by detecting and utilising social cues, facial recognition and emotions (Selcuk et al, 2024).

Social robots that function as AI agents capable of interactive behaviours and emotional expressiveness function as tutors, peers and even  teaching assistants. Belpaeme (2018) found that social robots can yield cognitive and affective outcomes comparable to human tutoring, thanks in part to their physical presence in the classroom. A similar study was conducted in a high school in 2025, where the humanoid robot Pepper, integrated with a large language model (ChatGPT), presented new learning content to students (Sievers, 2018). All participants found the delivery appropriate and the robot’s presence meaningful (Brender, 2021). Additionally, the AV1 robot enables students with chronic illness to virtually attend classes by absorbing class content and relaying it back to the student (The Times, 2023). This can significantly help students reintegrate into school life after absence.

Although there are a variety of benefits to the use of AI-enabled robots in education, there are some drawbacks that are important to consider including:

• Emotional Attachment Risks. Students, especially younger children, may form emotional attachments to the robots. The case of “Moxie”, the robot toy that shut down in 2024, showed that children felt grief and abandonment when the service ended (Prather, 2025).
• Dehumanisation of Learning. Critics argue that automating empathy through AI may encourage a transactional model of education (Creely & Blannin, 2024), where students are faced with programmed responses rather than nuanced, compassionate human judgement.

3.2 The impact and applications of AI-robotics in healthcare

Machine learning algorithms can analyse vast amounts of patient data, including medical records, imaging scans and genetic information, to identify patterns and make accurate diagnostic predictions. Robotic systems equipped with AI algorithms can assist healthcare professionals in interpreting diagnostic tests with greater speed and accuracy, reducing the risk of human error and misdiagnosis (De Togni et al., 2021). AI-driven robotics also enable personalised medicine by analysing individual patient data to tailor diagnostic and treatment approaches (Aymerich-Franch, 2020). By leveraging machine learning algorithms, robotic systems can identify unique biomarkers, genetic predispositions and treatment responses, allowing for precision medicine interventions. This personalised approach enhances diagnostic accuracy and therapeutic efficacy, ultimately improving patient outcomes and reducing healthcare costs (Braun & Clarke, 2008).

Similarly, the integration of Al and robotics in surgical interventions has revolutionised the field of medicine, enhancing precision, reducing human error and improving patient outcomes. AI-driven robotic systems offer submillimetre precision, surpassing the capabilities of human hands. This precision is particularly crucial in delicate procedures such as neurosurgery and ophthalmic surgery, where even minor errors can have severe consequences. Robotic-assisted surgeries facilitate minimally invasive techniques, reducing trauma, blood loss and recovery time for patients (Fitzgerald, 2014). 

Additionally, assistive robotics can aid advancements in drug discovery, accelerating drug screening and development. Robotics can automate the process of testing thousands to millions of compounds for their biological activity against a target, accelerating the screening process exponentially. Robotic systems integrated with microfluidic devices allow for miniaturised, high-throughput assays. These technologies enable rapid testing of compounds in small volumes, reducing reagent consumption and cost while increasing speed. AI algorithms can identify patterns and correlations in complex biological data, assisting in the discovery of potential drug candidates (Noble, 2018).

Overall, the influence of AI-powered robotics in healthcare is profound and offers a range of benefits that could revolutionise the system itself (Ness et al., 2024). These include:

• Decreasing Human Error. AI-powered robots can perform precise tasks that often surpass human capabilities (Ness et al., 2024). According to Ness et al. (2024), robots are able to execute procedures with minimal invasiveness and high accuracy, improving patient outcomes. Moreover, the digitisation of medical records provides vast amounts of data that AI can harness to improve diagnosis, prevent complications and reduce errors (Kabanda et al., 2024). AI helps detect complex conditions, address issues like inappropriate antibiotic prescriptions and identify potential mistakes in patient measurements (Kabanda et al., 2024).
• Offering Emotional Support. According to Ragno et al. (2023), social robots may be able to reduce anxiety, depression and other mental illnesses in fields such as psychiatry, paediatrics, geriatrics and rehabilitation. By providing therapy, personalised support and behavioural activation, AI-driven robots enhance patient wellbeing while reducing the workload of healthcare providers, enabling more consistent and accessible patient-centred care (Ragno et al., 2023).
• Reducing Workload. Social robots can reduce the workload for healthcare professionals by performing regular tasks, such as collecting vital signs and patient data, enabling clinicians to focus more on complex responsibilities (Ragno et al., 2023; Ness et al., 2024). Additionally, the use of robotic systems has been associated with improved workflow efficiency and enhanced coordination within medical teams, further optimising healthcare delivery (Ness et al., 2024).

Despite its potential, the integration of AI and robotics into healthcare also presents significant challenges. Resistance to technology among healthcare professionals is common, stemming from concerns over job displacement, privacy issues and bias (Ness et al., 2024). These issues can slow adoption rates and create friction in implementing new technologies.

• Job Displacement. AI’s ability to automate tasks traditionally performed by humans may lead to reduced demand for certain medical roles. For example, since AI can process and analyse images and data efficiently in a fraction of a human’s time, radiologists face the prospect of reduced salary or future unemployment (Pham et al., 2024).
• Privacy Threats. Privacy and data security remain paramount concerns as AI relies on large datasets that may be vulnerable to breaches or misuse, potentially jeopardising patient confidentiality (Ness et al., 2024). Furthermore, many AI systems are managed by private companies with limited transparency and oversight, increasing the risk of unauthorised access or exploitation of sensitive data (Murdoch, 2021).
• Bias. Algorithmic bias poses a real risk, with AI tools discriminating against underrepresented groups if training data lacks diversity, subsequently exacerbating healthcare inequities (Ragno et al., 2023; Ali, 2023). Bias can seep in at multiple stages, resulting in skewed clinical outcomes that disproportionately affect marginalised populations (Nazer et al., 2023). Nazer et al. (2023) note that AI is predominantly trained on data frim the US or China, thereby limiting model performance across different regions and demographic groups.

4. Ethical and Social Challenges of AI-Enhanced Robots

As AI-enhanced robots become more prominent in day-to-day life, complex ethical challenges emerge. These include concerns about bias, fairness, transparency, privacy, consent, human dignity, accountability and autonomy. The prompt advancements in AI, machine learning and human-robot interaction enables robots to act independently and form emotionally engaged relationships, raising questions of overreliance and social displacement.

4.1 Bias, Fairness and Transparency

Grover (2025) highlights dilemmas such as algorithm bias, accountability fairness and governance, calling for transparent policies and interdisciplinary oversight. Similarly, Benjamin (2025) emphasises that as robots gain increasing autonomy, issues around fairness, transparency and safety become central concerns. The UNESCO framework on AI ethics further underscores principles such as transparency, privacy, accountability, explainability fairness and human oversight, advocating for explainable methods that use standard metrics while maintaining user-friendly interfaces to foster accountability, accessibility and trust (AHEG, 2021).

4.2 Consent, Privacy and Human Dignity

Social robots in sensitive contexts, such as mental health support, raise acute ethical concerns. Research shows that their use can inadvertently compromise privacy, informed consent and the quality risk of human care. For instance, studies on “PARO” and “LOVOT” robots (see Section 5) bring attention to the potential risks of social robots, including the replacement of human caregivers as well as the infantilisation of the patient (Hung et al., 2025; Leinweber et al., 2025). Moreover, Leinweber (2025) conducted a systematic review of social robotics in elderly care and identified over 60 ethical aspects, ranging from unresolved challenges all the way to hazards, emphasising the need for context-specific decision making. These findings further suggest that while social robots are able to provide a companion, their integration must prioritise dignity, equity and consent. 

Privacy is also a significant central concern. Vozna and Costantini (2025) state that AI-driven robots in healthcare must comply with data protection laws such as the General Data Protection Regulation (GDPR) and the Personal Data Protection Law (PDPL) to ensure data is stored securely and to prevent the misuse of sensitive biometric information.

Additionally, Benjamin (2025) points out that as robots act more independently, it becomes unclear as to whether liability should fall on manufacturers, developers or even end-users. Transparency and auditability are essential to avoid the risk of unaccountable “black box” models.

In military contexts, lethal autonomous weapons raise further alarms. Austria has pushed urgent regulations of such systems, stressing the need for human control over life-and-death situations and decisions (United Nations, 2023). Many critics also argue that fully autonomous weapons may breach humanitarian principles of distinction and proportionality (Müller et al., 2024). 

5. The Future of Human-Robot Relationships

Human-robot relationships (HRR) are emerging as a significant field of study, drawing a range of majors from psychology, engineering and social sciences to understand how people interact with increasingly intelligent machines. With integration of attachment theory, researchers have been able to begin to explain how humans can form genuine emotional bonds with robots, experiencing companionship, trust and improved mental health. Simultaneously, there have been rapid advances in artificial intelligence (AI), machine learning (ML) and biometric technologies which have transformed robots from simple mechanical tools into emotionally sympathetic, empathetic partners capable of adapting to the user’s needs. Such developments allow exciting possibilities to be created, further enhancing companionship, collaboration and social support, particularly for vulnerable populations. However, the future of HRR is also shaped by boundaries and concerns, including issues with trust, ethics, privacy and safety which will define how far HRR relationships are able to progress and how society plans to integrate these robots into everyday life.

5.1 Companionship and bonding

HRR have had an increasing amount of testing using attachment theory, providing explanations for how people bond with entities that are not human. Laban et al. (2024) found that during an in-depth, longitudinal study, participants gradually divulged personal information to robots, giving more sensitive information over time, which correlated with a reduction in loneliness and a significant improvement in subjective wellbeing (Laban et al., 2024). Disclosure was used as an indicator of relational depth, insinuating that robots can function as a partner over extended periods of time with human interaction. Saito and Yamamoto (2024) extended the finding in a practical context by studying the LOVOT robot with adults who were not in a romantic relationship (see Figure 1).

Figure 1: Japanese LOVOT Robots (Tan et al., 2024).

Saito and Yamamoto’s research showed a significant benefit in discerned social support and a decrease in seclusion, further confirming that robots are able to provide stable companionship gains in everyday life. Broadbent (2021) argued that human-animal relationships may provide a suitable analogy for HRI than human-to-human relation, considering attachment can happen in asymmetrical ways, where one party lacks the complex cognition. The relative chassis suggests that expectations of exchange are not always required for emotional bonds to be able to form (Zlotowski et al., 2021). This also emphasises that applying attachment theory to HRI allows researchers to better explain mechanisms by which users begin to perceive robots as associates other than tools (Zlotowski et al., 2021). As a result, these findings further suggest that robots can take on roles that extend beyond instrumental functions and begin to meet emotional and social needs, particularly for vulnerable populations such as children, the elderly and socially isolated individuals.  

5.2 Technology facilitators of HRI

Advances in AI and machine learning are the foundation of progressively refined human-robot relationships. Chen and Lee (2025) observed that natural language processing (NLP) and affective computing allows robots to engage in human-like conversations, which enables users to comprehend robots as sympathetic, empathetic and trustworthy partners. Indeed, the focus of a social robot design has shifted from mechanical functionality (movement) to affective engagement, which reflects the growing recognition of emotional communication as central to HRI (Ahmed et al., 2024). Mortensen (2025) confirmed this through a systematic literature review, highlighting that interdisciplinary approaches are now standard in human-robot relationships research, with engineers working alongside psychologists, sociologists and robot ethicists to develop robots with human-like psychological and ethical capacities. The capacity of robots to adapt their behaviour is particularly important; by analysing and comparing the response of the user with their history of data, robots are able to adjust and personalise their responses (García-Martínez et al., 2024). This capability of AI adaptations reduces the need for manual programming, allowing robots to refine their companionship-like behaviours over time. 

Biometric technologies, such as speech recognition, significantly facilitate HRI, allowing robots to process a wide variety of languages and human accents (Soori et al., 2023). Guo et al. (2024) further explore physiological biometrics which allow the robot to analyse human emotions from a variety of facial features, ranging from the tips of the eyebrows to the corners of the mouths. By allowing these AI-driven advancements to happen, we are permitting robots to be more contextually aware, responsive and socially engaging, giving the user the appropriate response. Collectively, this strengthens the possibility of long-term relationships and bonds between humans and robots by providing interaction experiences that are progressively personalised and emotionally convincing, deepening the internal trust and bond in HRR.

6. Conclusion

The rise of AI-enabled robotics has redefined the societal roles assigned to robots. In education, social robots enhance personalised learning in classrooms and encourage engagement in STEM disciplines. In healthcare, robots ameliorate diagnostic accuracy, improve surgical precision and offer emotional support, subsequently alleviating pressures on medical professionals and improving patient outcomes.

Despite the benefits, AI-enhanced robots have introduced many ethical and social challenges. Transparency, privacy, fairness and human dignity continue to be concerns, with UNESCO highlighting the necessity of explainability, oversight and equity to build trust in human-robot systems. Questions continue to be raised on the liability and responsibility of robotos, emphasising the importance of maintaining human control for ethical decision-making.

By embedding principles of human oversight, transparency and fairness in the design of these autonomous robots, they become a tool rather than a replacement, upholding human dignity, values and ethics.

Ad Hoc Expert Group (AHEG) for the Preparation of a Draft text of a Recommendation on the Ethics of Artificial Intelligence (2020) Outcome document, first draft of the Recommendation on the Ethics of Artificial Intelligence, UNESCO [online]. <https://unesdoc.unesco.org/ark:/48223/pf0000373434>

Ahmed, E., Buruk, O. & Hamari, J. (2024) Human-Robot Companionship: Current Trends and Future Agenda, International Journal of Social Robotics, 16, pp. 1809-1860.

Ali, M. (2023) A Comprehensive Review of AI’s Impact on Healthcare: Revolutionizing Diagnostics and Patient Care, Jurnal Multidisiplin Ilmu, 4(4), pp.1163-1173.

Aymerich-Franch, L. (2020) Why it is time to stop ostracizing social robots, Nature Machine Intelligence, 2(364).

Belpaeme, T., Kennedy, J., Ramachandran, A., Scassellati, B. & Tanaka, F. (2018) Social robots for education: A review, Science Robotics, 3(21).

Bishop, L., van Maris, A., Dogramadzi, S. & Zook, N. (2019) Social robots: The influence of human and robot characteristics on acceptance, Paladyn Journal of Behavioral Robotics, 10(1), pp. 346-358.

Braun, V. & Clarke, V. (2008) Using thematic analysis in psychology, Qualitative Research in Psychology, 3(2), pp. 77-101.

Brender, J., El-Hamamsy, L., Bruno, B., et al. (2021) Investigating the Role of Educational Robotics in Formal Mathematics Education: The Case of Geometry for 15-Year-Old Students, Technology Enhanced-Learning for a Free, Safe, and Sustainable World [conference paper], pp. 67-81.

Broadbent, E. (2016) Interactions With Robots: The Truths We Reveal About Ourselves, Annual Review of Psychology, 68, pp. 627-652.

Coeckelbergh, M. (2011) Can we trust robots? Ethics and Information Technology, 14, pp. 53-60.

Creely, E. & Blannin, J. (2025) Creative partnerships with generative AI. Possibilities for education and beyond, Thinking Skills and Creativity, 56(101727).

Cyberne1 (2015) 1968 Minsky-Bennett Arm – Marvin Minsky and Bill Bennett (American), Cyberneticzoo [online]. <https://cyberneticzoo.com/underwater-robotics/1968-minsky-bennett-arm-marvin-minsky-and-bill-bennett-american/>

De Togni, G., Erikainen, S., Chan, S. & Cunningham-Burley, S. (2021) What makes AI “intelligent” and “caring”? Exploring affect and relationality across three sites of intelligence and care, Social Science & Medicine, 277(113874).

Farouk, M. (2022) Studying Human Robot Interaction and Its Characteristics, International Journal of Computations, Information and Manufacturing, 2(1).

Gruen, L. & Monso, S. (2024) The Moral Status of Animals, Stanford Encyclopedia of Philosophy [online]. <https://plato.stanford.edu/entries/moral-animal/>

Guo, T., Obidat, O., Rodriguez, L., Parron, J. & Wang, W. (2022) Reasoning the Trust of Humans in Robots through Physiological Biometrics in Human-Robot Collaborative Contexts, 2022 IEEE MIT Undergraduate Research Technology Conference (URTC).

Kabanda, N. (2024) The Impact of AI on Reducing Medical Errors, Kiu Publication Extension [online]. <https://www.researchgate.net/publication/384455361_The_Impact_of_AI_on_Reducing_Medical_Errors>

Kaur, J., Kamal, R., Kaur, M. & Malhan, S. (2024) The Symphony of Efficiency: Transforming Healthcare Through Robotic Process Automation, in Tripathy, R.K., Bhattacharya, P. & Liu, H. (eds) Clinical Practice and Unmet Challenges in AI-Enhanced Healthcare Systems, pp. 245-261.

Ketterl, M., Jost, B., Leimbach, T. & Budde, R. (2015) Open Roberta – a Web-Based Approach to Visually Program Real Educational Robots, Læring & Medier, 14 [online]. <https://tidsskrift.dk/index.php/lom/article/view/22183/20134>

Laban, G., Kappas, A., Morrison, V. & Cross, E.S. (2023) Building Long-Term Human-Robot Relationships: Examining Disclosure, Perception and Well-Being Across Time, International Journal of Social Robotics, 16, pp. 1-27.

Mohammed, A. (2024) Advanced Guide for Natural Language Processing, Analytics Vidhya [online]. <https://www.analyticsvidhya.com/blog/2023/09/advanced-natural-language-processing-nlp/>

Murdoch, B. (2021) Privacy and artificial intelligence: challenges for protecting health information in a new era, BMC Medical Ethics, 22(122).

Ness, S., Xuan, T.R. & Oguntibeju, O.O. (2024) Influence of AI: Robotics in Healthcare, Asian Journal of Research in Computer Science, 17(5), pp. 222-237.

Noble, S.U. (2018) Algorithms of oppression: How search engines reinforce racism, New York University Press.

Onyeulo, E.B. & Gandhi, V. (2020) What Makes a Social Robot Good at Interacting with Humans?, Information, 11(1), 43.

Ouyang, F. & Xu, W. (2024) The effects of educational robots in STEM education: a multilevel meta-analysis, International Journal of STEM Education, 11(7).

Oxford English Dictionary (n.d.) Robot, OED [online]. <https://www.oed.com/dictionary/robot_n2?tl=true&tab=factsheet>

Pham, T.B., Ismayilova, A., & Gutierrez, L.P. (2024) Is AI Causing The Jobs Diplacement In Medical Industry?, Displacement (Psychology) [online]. <https://www.researchgate.net/publication/380210735_Is_AI_Causing_The_Jobs_Displacement_In_Medical_Industry>

Prather, A. (2025) What Happens When Robots Die? The Case for End-of-Life Standards in Robotics, Six Degress of Robotics [online]. <https://sixdegreesofrobotics.substack.com/p/what-happens-when-robots-die-the>

Prescott, T.J. & Robillard, J.M. (2020) Are friends electric? The benefits and risks of human-robot relationships, iScience, 24(1), 101993.

Ragno, L., Borboni, A., Vannetti, F., Amici, C. & Cusano, N. (2023) Application of Social Robots in Healthcare: Review on Characteristics, Requirements, Technical Solutions, Sensors, 23(15), 6820.

Selcuk, N.A., Kucuk, S., & Sisman, B. (2024) Does really educational robotics improve secondary school students’ course motivation, achievement and attitude?, Education and Information Technologies, 29, pp. 23753-23780.

Sievers, T. (2025) A Humanoid Social Robot as a Teaching Assistant in the Classroom, arXiv, 2508.0564.

Sundar, S.S. (2020) Rise of Machine Agency: A Framework for Studying the Psychology of Human-AI Interaction (HAII), Journal of Computer-Mediated Communication, 25(1), pp. 74-88.

The Times (2013) The robots letting pupils with cancer learn from hospital, The Times [online]. <https://www.thetimes.com/uk/education/article/meet-the-robots-letting-pupils-with-cancer-learn-from-hospital-lpxcb0v07>

Understanding Empire (2014) The Laser Age in the Vietnam War, Understanding Empire: Technology, Power, Politics [online]. <https://understandingempire.wordpress.com/tag/laser-guided-bomb/>

Walters, M. (n.d.) Human-Robot Manipulation and Co-Operation [pdf]. <https://www.herts.ac.uk/__data/assets/pdf_file/0004/61609/HumanRobotCooperation.pdf>

Willmer, G. (2025) Robo-companion: humanoid robot gets chatty to help elderly hospital patients, Horizon: The EU Research & Innovation Magazine, [online]. <https://projects.research-and-innovation.ec.europa.eu/en/horizon-magazine/robo-companion-humanoid-robot-gets-chatty-help-elderly-hospital-patients>

Zahari, M.Z.A.M. (2019) Moving beyond Protection with PULSE – Using A.I. to make healthcare more affordable and accessible, International Journal of Human and Health Sciences [online]. <https://ijhhsfimaweb.info/index.php/IJHHS/article/view/134>