Ontario secondary school leaders navigate GenAI integration with competing perspectives—viewing it both as an opportunity for enhanced productivity and personalized learning and as a threat to academic integrity and authentic student voice—while facing significant equity gaps, capacity limitations, and an absence of clear provincial policy guidance.
Objective
The primary objective of this exploratory study was to examine how Ontario secondary school leaders understand and view the integration or non-integration of generative artificial intelligence (GenAI) in their schools, including their rationale, motivations, and perceived barriers and supports. The study aimed to address a significant research gap, as most GenAI studies focus on higher education and international contexts rather than Ontario's K-12 system. With over two million students served by Ontario schools and a recently updated high school curriculum requiring technological education credits, yet no provincial policies regarding GenAI use, this research sought to capture secondary school leaders' perspectives during Fall 2024. The study explored two research questions: (1) How do Ontario secondary school leaders understand the integration of GenAI in their schools, and what values inform their perspectives? (2) What do Ontario secondary school leaders perceive as the key opportunities and challenges of integrating GenAI in their schools?
Methods
This qualitative study employed a basic qualitative research design utilizing reflexive thematic analysis grounded in e-leadership theory, pedagogical beliefs, and educational change theory. The researcher conducted semi-structured 60-minute interviews via Microsoft Teams with 20 secondary school leaders from 11 different school boards across Ontario during Fall 2024. The sample included department heads (n=14) from seven subject areas, vice principals (n=3), one system principal, and two other school leaders. Participants represented rural (n=4), suburban (n=5), urban (n=8), and mixed geographic contexts (n=3), with 18 from English public boards and 2 from English Catholic boards.
Recruitment occurred through posters distributed via online forums, mailing lists, social media (particularly LinkedIn), and professional networks. After initial limited response, the researcher amended the ethics protocol to offer a $20 gift card as compensation. Data analysis followed Braun and Clarke's (2006) six-step reflexive thematic analysis process, resulting in 166 distinct codes organized into five major themes. To enhance credibility, the researcher engaged in peer debriefing, shared findings with the dissertation supervisor, and conducted member-checking with participants.
Key Findings
Five major themes captured Ontario secondary school leaders' complex perspectives on GenAI integration. First, motivations for integration included viewing GenAI as valuable for enhancing productivity (email writing, report card comments, lesson planning, reference letters), reducing cognitive load, addressing teacher shortages, and preparing students for AI-enabled futures. Leaders displayed technocratic assumptions about efficiency and believed GenAI could improve work-life balance. Personal characteristics such as openness to innovation influenced willingness to integrate. Leaders demonstrated e-leadership by modeling ethical GenAI use, though integration timing varied dramatically—some began immediately after ChatGPT's release while others remained hesitant into 2024.
Second, concerns about GenAI centered on it hindering student learning and undermining educators' responsibilities. Fifteen of twenty participants voiced concerns about GenAI as a copyright issue and threat to creative capacity, with some viewing its use as a breach of professional responsibility. The predominant initial association was fear of student cheating and academic dishonesty. Leaders particularly worried about students losing critical skills (writing, critical thinking, creativity) and authentic student voice—especially in English, social sciences, and humanities. Leaders emphasized their role in preserving diverse student voices, particularly for multilingual learners, arguing that GenAI homogenizes expression toward dominant (White, middle-class, educated) voices. Data privacy concerns included protecting student information from corporate intrusion. Some leaders believed there was "no learning curve" to GenAI and that their role was to teach foundational skills directly rather than through AI mediation.
Third, opportunities and challenges for equitable learning revealed tensions. Opportunities included personalized learning (lesson plans for students with disabilities, materials in students' home languages), leveling the playing field by providing tutor-like support for students who cannot afford private tutoring (particularly valuable for rural students), scaffolding learning, and enhanced learning across various streams. However, leaders noted university-preparation students appeared more likely to use GenAI ethically while college-preparation students were more often accused of plagiarism, potentially reflecting classist/racist biases. Challenges included changing assessment practices with mixed approaches (application-based assessments, pen-and-paper work, emphasis on process over product, surveillance tactics), unreliable plagiarism detection tools, inconsistent disciplinary approaches, and questions about citing ChatGPT. Leaders across experience levels expressed resistance due to bias in GenAI (Eurocentric, disadvantaging multilingual learners, reproducing colonialism), inaccuracy, established teaching practices that already work, and concerns that GenAI cannot replicate authentic teacher-student relationships.
Fourth, significant equity challenges persisted regarding access, connectivity, and cost. Access variation occurred across Ontario: some schools blocked GenAI tools while others permitted full access. Device access varied dramatically—some schools provided one-to-one Chromebook programs with take-home options while others had only five Chromebooks for classes of 32 students. Rural areas, lower socioeconomic communities, and newcomer families faced particular barriers to home internet access and devices. Using advanced GenAI features requiring paid subscriptions created sharp distinctions between private and public schools. Capacity limitations included lack of time given heavy workloads, absence of clear board-level and provincial direction, limited professional development (mostly one-time workshops), and concerns about student maturity and capacity to use technology effectively.
Fifth, leaders called for comprehensive GenAI policy involving multiple stakeholders. Recommendations included acknowledging GenAI's existence and embedding it throughout educational domains, establishing parameters for safe use, combining top-down direction with grassroots input, recognizing that "one size does not fit all," considering age restrictions, investing in equitable technology access, providing ongoing tailored professional development, updating curriculum and assessment practices, establishing stricter plagiarism policies for both educators and students, proceeding with cautious integration informed by past technology initiatives, or alternatively, not integrating GenAI while focusing resources on fundamental needs like breakfast programs. Leaders identified necessary stakeholders as teachers, parents, students, Ministry of Education, unbiased experts, administrators, community members, and various school personnel.
Regarding tools, leaders reported using ChatGPT, Microsoft Copilot, Google Gemini, and numerous other platforms. Only three tools received board approval: Adobe Express, Google Gemini, and Microsoft Copilot—though boards provided approved tools without instructional support. Parental and community engagement appeared minimal, with most insights coming from leaders' assumptions rather than direct conversations.
Implications
This research reveals critical implications for navigating educational technology integration. For students, learning experiences with GenAI vary dramatically depending on location, socioeconomic status, board policies, individual teacher beliefs, and academic stream—creating significant knowledge inequities. For teachers and school leaders, AI literacy cannot be separated from questions of professional identity, workload, capacity, and values. Professional development must go beyond technical training to engage educators in critical reflection about their roles and how GenAI challenges or supports educational purposes. Teacher resistance often stems from legitimate concerns about authenticity, equity, student voice, data privacy, and preserving human relationships rather than technophobia.
For administrators and policymakers, the findings reveal a system in flux with significant gaps between technological change and institutional response. The absence of provincial policy creates inconsistent board-level decisions, leaving educators without clear guidance. However, premature, overly prescriptive policies could stifle productive experimentation and fail to account for contextual differences. Critical equity implications emerge regarding access disparities—some schools provide one-to-one devices while others have five Chromebooks for 32 students, revealing Ontario's system remains fundamentally unprepared for equitable GenAI integration. Before mandating GenAI use, the province must address foundational infrastructure inequities in devices, connectivity, and technical support.
Limitations
The study's limitations include its exploratory, cross-sectional nature capturing perspectives only during Fall 2024 during rapid technological and policy evolution. The sample lacked principals (except one system principal) and focused only on public schools in Ontario, limiting generalizability beyond this context. Self-reported data may not reflect actual practices or future perspectives. The study did not collect data on participants' intersectional identities (race, class, gender beyond leadership role), which may influence perspectives. Sample recruitment through professional networks and LinkedIn may have attracted participants already engaged with educational technology, potentially biasing toward more tech-positive perspectives. The researcher's outsider position to secondary school leadership required careful reflexivity to avoid misinterpreting institutional contexts and power dynamics.
Future Directions
Future research should conduct longitudinal studies tracking how perspectives, policies, and practices evolve as provincial and board-level policies emerge. Comparative studies examining public versus private schools, different provinces/countries, and various disciplines could illuminate contextual factors shaping integration. Research should investigate whether streaming differences in GenAI use reflect real patterns or teacher biases, include principal perspectives missing from this study, examine student and parent perspectives directly rather than through leader accounts, test effectiveness of different professional development approaches, and explore how GenAI integration intersects with other educational priorities like mental health, equity, and fundamental skill development.
Title and Authors
Title: "Rise of Generative Artificial Intelligence: Insights from Secondary School Leaders"
Author: Fung Yu Nancy Hsiung (University of Toronto)
Committee: Dr. Caroline (Carly) Manion (Chair), Dr. LaVette Burnette (Co-Chair), Dr. J. Charlene Davis (Member)
Published On: October 2025 (Master of Arts degree awarded)
Published By: Master's thesis submitted to the Department of Leadership, Higher and Adult Education, Ontario Institute for Studies in Education, University of Toronto