Schools, as institutions for learning, should be safe places for students to learn and socialize. However, it’s undeniable that violence remains a frequent issue, both physical and verbal. Behaviours such as bullying, physical fights, and verbal and nonverbal harassment are triggers for increased anxiety, depression, and mental health issues in students ( Arslan et al., 2021; Kaspar, 2013; Kim et al., 2020; Liu et al., 2025; Wang et al., 2024). In addition, this form of violence also causes psycho-social and academic problems for students at school and at home ( Gutzwiller-Helfenfinger, 2021; Kernbach-Wighton, 2022). In recent years, attention to this issue has increased as public awareness of the importance of creating a safe, comfortable, and healthy learning environment for students has grown. However, prevention efforts are often reactive and suboptimal.

Therefore, schools need to manage their student counseling services wisely. Today, technological advances, particularly artificial intelligence, help monitor and analyze student behavior in the classroom, identifying disciplinary violations and bullying early ( Sidhu & Sidhu, 2025; Zheng et al., 2025). Schools have begun adopting technologies such as counseling chatbots and automated notification systems via WhatsApp or SMS. AI can significantly improve mental health services in schools by providing real-time physiological monitoring, emotional regulation, and cognitive behavioral modification ( Guo & Hou, 2021).

AI chatbots can offer immediate, non-judgmental feedback and support, although they should be complemented by human counselors to address issues of empathy and trust ( Al-Shraifin et al., 2025; Chan, 2025). AI should be used to complement and enhance human interaction, rather than replace it. This approach can improve teacher-student relationships, reduce anxiety, and increase students’ willingness to communicate ( Guan et al., 2025). However, the integration of AI in managing student welfare in Indonesia remains very limited. Many schools have not yet integrated AI into their administrative systems for student counseling services, and therefore, the technology’s benefits in violence prevention have not been fully realized.

This is where the quality of counseling administration comes into play. Collaborating with the school isn’t just about student data collection, but also about how well the school can process and manage information and data, collaborate with teachers, counselors, and the principal, and practice case management ( Chuang & Liu, 2022; Li-Min & Hsiu-Hao, 2022; Lowery et al., 2020; Solmonson, 2024; Yordy et al., 2022). This component contributes to creating a healthy mental environment and prioritizing psychological affirmations to support students’ academic achievement. The success of school counseling administration when integrated with AI will help detect and monitor trustworthy data to create healthy schools. It also provides early protection from risky behaviors that can lead to unwanted negative behaviors.

Furthermore, the successful implementation of AI technology and the administrative management of school counseling will not be successful without the support of the school organization. Such support can come in the form of policies and commitments from the principal to encourage the use and provision of digital counseling services for students to increase accessibility and effectiveness ( Çela et al., 2025; Madeline, 2025). Schools with collaborative and well-being-focused cultures will seek to integrate AI into counseling and violence prevention services. This aims to enhance services aimed at addressing student mental health crises by providing supportive real-time guidance and early identification of mental health issues ( Al-Shraifin et al., 2025; Aziz et al., 2025; Baksi et al., 2025; Guo & Hou, 2021), thus creating a healthy and safe school environment. Therefore, the combination of AI, school counseling administration, and organizational support forms a solid, interdependent unit.

Figure 1 illustrates the bibliometric network visualization generated using VOSviewer for publications over the last five years, research on the use of AI in education has continued to grow. Based on a bibliometric analysis using VOSviewer, data covering 17,660 articles and keywords such as AI, school counseling services, school violence, students, education, bullying, and mental health, the analysis shows that the yellow cluster surrounding the keywords students, education, psychology, and adolescents indicates that in the past two years (2023–2024), there has been increased attention, but research remains limited. However, one interesting aspect of the mapping results is the absence of keywords related to school violence, early risk detection, school counseling, and AI. This indicates that while there is a wealth of research related to AI and education, there is little research on AI for school counseling service management systems. Thus, the VOSviewer results highlight a research gap that remains largely unexplored: how schools use AI as a tool to support counseling services to detect risks and prevent violence in schools.

This study has provided more detailed information related to gaps or branches in bibliometric research as presented in the VOSviewer map above. This study has integrated AI along with the quality of administrative services in school counseling and school organizational support to see how the three interact in improving risk identification and early violence prevention in Indonesian secondary schools. Establishing AI as a variable that is more than just a technological tool but as a component of the counseling service system along with school administrative leadership that supports organizational policies. This is rare in international research. The context of secondary schools in Indonesia has not been widely explored as shown in the VOSviewer analysis, particularly regarding student protection, counseling, and violence prevention. This study is expected to answer how AI can be a valuable resource in preventing violence in schools in real terms and not only on paper or as a conceptual framework. This research can add to the global discourse regarding the use of AI in the administrative practices of school counseling services and student protection policies in secondary schools in Indonesia.

The greatest impact of AI on the quality of counseling administration (β = 0.586) indicates that technologies such as AI have significantly improved the efficiency of counseling services. This finding aligns with previous research, where AI plays a role in analyzing and diagnosing data and discovering patterns missed in traditional analysis ( Garcia, 2025; Sun et al., 2025). In addition, AI in the case management system is also useful in helping to schedule consultation sessions and reminders, thereby reducing the significant administrative burden in the process of implementing school counseling services ( Yu Wu et al., 2025). Technology acts as a powerful facilitator in improving the accuracy, efficiency, and integration of counseling service data ( Muhammad, 2024). By leveraging advanced AI-based data analytics, digital tools, and comprehensive data systems, counselors can provide more accurate, efficient, and accessible services.

On the other hand, the direct effect of artificial intelligence on early risk detection was smaller, at β = 0.163. This finding reinforces the view in the educational technology literature that the effectiveness of AI-based systems is highly dependent on the implementation context and organizational readiness. Successful AI implementation requires a robust technology infrastructure and compatibility with existing systems. Challenges include integration with legacy systems, data quality issues, and the complexity of AI systems ( Madanchian et al., 2025; N. M. et al., 2025; Pathak & Bansal, 2025). In this context, the mediating role of the quality of counseling administration and significant organizational support reinforces socio-technical systems theory, which asserts that the recursive interaction between technology and people, within the context of social practices and norms, is crucial for effectively utilizing IT resources. This perspective helps in understanding how organizational and technological resources interact to improve performance ( Kaur et al., 2024). The large coefficient of organizational support’s influence on early risk detection (β = 0.497) indicates that supportive policies, leadership, and organizational culture are instrumental in enabling AI analysis to be transformed into preventive action. This is in line with Sulaymani et al. (2022) who stated that before intelligent technology can be adopted in educational institutions, structural support is needed so that this intelligent technology can contribute to strategic decision-making ( Sulaymani et al., 2022).

The significant mediating role of administrative quality of counseling and organizational support in this model supports the strengthening of the socio-technical systems theory approach, which states that technological performance is a product of the interaction between technical and social components in an organization ( Bali Kamra Professor et al., 2024; Davis et al., 2025). The large coefficient of influence of organizational support on early risk detection (β = 0.497) indicates that organizational support is critical as a form of analysis of the output of the AI analytical model system into a useful preventive measure. This is in line with the results of previous research that the adoption of intelligent technology in educational institutions requires structural support to contribute to strategic decision-making ( George & Wooden, 2023; Rico-Bautista et al., 2020).

The R ² coefficient value of 0.518 means that 51.8% of the total change in early risk detection of school violence can be explained by a combination of variables: artificial intelligence, quality of counseling administration, and organizational support. These results indicate that the model has moderate explanatory power. This further confirms that the approach taken is more integrative than purely deterministic models that rely solely on technology. This also provides a positive contribution to previous research that has only placed AI in a dominant position and has not explicitly recognized the contribution of organizational support in early detection of school violence.

Overall, this research supports the view that the strategic value of applying AI to educational administration lies in its ability to improve administrative efficiency and early risk detection for school violence prevention. AI drives decision-making through proactive risk management, thereby leading to a more resilient education ecosystem. The positive impact of AI in the education system is undeniable, and the need for ethical considerations and practical challenges to address this need to be addressed ( Kilianova et al., 2025; Mitchell, 2025; Pavunraj et al., 2025).

AI performs well in detecting early risks of violence in secondary schools, particularly when coupled with a counseling service administration system and organizational support from the school. The use of AI has the potential to detect risks earlier, with the greatest potential for AI to function within a more structured administrative system, supported by relevant school policies, leadership, and resources. This finding is considered to almost confirm that the risks detected are not solely due to technology, but rather to organizational systems that need to be integrated and aligned with technology and administrative systems. Furthermore, in the context of this research, empirically demonstrates that an integrative model connecting AI, the quality of counseling service administration, and organizational support can provide a deeper understanding of how schools function, through a reactive approach, towards proactive, data-driven violence prevention and student protection.

These findings provide practical and theoretical contributions, particularly in the context of the school counseling system in Indonesia, and support the need for responsible use of AI in educational administration. Essentially, this research confirms that to continuously detect risks and prevent violence in schools, relying not only on technology but also on the extent to which school organizations manage, understand, and utilize technology to protect and prioritize student well-being. This study strengthens the socio-technical systems theory regarding the use of AI in supporting early identification of violence risks in schools, however, its effectiveness depends on the administrative quality of guidance and counseling services and organizational support in the form of leadership and policies. The key to success is that technology will only have an impact if its management is combined with a good administrative system. In addition, this study adds insight into the field of educational administration by positioning the administrative quality of guidance and counseling services as a strategic administrative mechanism that bridges the impact of the use of technological innovation and protection from violence against students.

Practically, the results of this study strengthen the quality of administrative services and organizational support for improving school counseling services by adopting the use of AI for early detection of the risk of violence against students. Schools need to develop better counseling systems, focus on case management, and facilitate the provision of ethical and sustainable resources. Furthermore, support in the form of teacher and counselor training and the integration of AI into the counseling service administration system to strengthen detection and enhance effective risk prevention needs to be enhanced.

The Research Ethics Committee of the Indonesian University of Education granted permission for this study before data collection (Approval Number: 072/ UN.40.EC/KP.09.03/2026). This study was conducted in accordance with the research guidelines of the ethics body and international ethical standards for research involving humans.

Data collection involved respondents aged 13–18 years. Therefore, written informed consent was obtained from the participants’ parents or legal guardians before they participated in the study. Permission to conduct the study was also granted by the relevant school institution. All participants were given an explanation of the study’s purpose and provided written informed consent before completing the questionnaire. Participants’ responses were collected anonymously and used for research purposes.