District Leaders Navigate AI Governance and Equity on Uneven Ground

Skip local navigation

Sub-Navigation Menu

District Leaders Navigate AI Governance and Equity on Uneven Ground
A Research Brief on AI Implementation

By Elizabeth Davis, Seth B. Hunter, and Megan O’Brien

June 2026

Districts are being inundated with promises of the power of AI to close learning gaps through personalized learning and tutoring, as well as to improve efficiency by supporting teachers with administrative tasks.i However, AI adoption is occurring with often contradictory state and federal policies, creating unclear compliance conditions for districts to navigate. At the same time, districts are under pressure from technology vendors to purchase AI software or upgrade systems to include AI so they do not fall behind other districts. This means that districts are often pushed into rapid adoption, while policies lag behind, struggling to keep pace with fast-paced technological change.ii If the promised advancements from AI use are to scale without contributing to inequities in education, districts must develop AI implementation strategies that consider ethical implications in a rapidly evolving landscape. In doing so, districts can determine if and how they want to proceed with AI adoption. Using qualitative data from interviews with district leaders, we examine the key considerations that inform decision-making around AI governance and equity and propose recommendations for district and state leaders based on our findings.

District Level AI Governance and Equity Concerns

District leaders are making AI adoption decisions in a policy environment that remains uneven and rapidly evolving. National evidence suggests that AI use in schools is expanding more quickly than clear policy guidance is developing. Research indicates that more districts are training teachers on AI and recognize that implementation requires system-level support rather than isolated, one-off uses by teachers.iii Major guidance documents reinforce this tension. The U.S. Department of Education’s Office of Educational Technology (2023)iv described AI as a potential support for teaching and learning, emphasizing privacy, bias, transparency, and human oversight, while UNESCO (2023)v similarly called for human-centered AI adoption that combines safeguards with long-term capacity building.

Ethical and equity concerns are inseparable from implementation. Emerging scholarship identifies recurring risks associated with AI adoption in education, including privacy and data protection, algorithmic bias and disparate impact, opacity and limited explainability, challenges with vendor transparency and accountability, and the reshaping of professional labor conditions.vi In K-12 settings, these concerns are particularly consequential because district decisions about AI adoption affect minors, intersect with legal and policy obligations (e.g., FERPA), and may unevenly shape access to educational opportunities across schools and communities. Ethical issues are not confined to classroom practice; they are embedded in district-level decisions regarding procurement, tool approval, data governance, acceptable use, and professional learning. Cheah et al. (2025)vii further found that teachers often remain underprepared to integrate generative AI and tend to use these tools more for planning and administrative tasks than for direct instruction, suggesting that limited training and weak policy guidance may intensify uneven implementation.

Furthermore, AI is being adopted in K–12 education within a system still marked by longstanding inequities in access and organizational capacity. Although AI is often promoted as a tool that can personalize learning and improve efficiency, its benefits are unlikely to be evenly distributed when the digital divide persists across schools and communities. Students in low-income communities and communities of color continue to experience less consistent access to reliable internet, devices, and technology-enabled learning opportunities.viii These conditions matter because AI adoption does not occur on neutral ground. Pham et al. (2024)ix argued that AI may deepen racial disparities in education if implementation proceeds without attention to structural inequality, access, and institutional readiness. Historically under-resourced schools may also lack the staffing and professional learning needed to use AI responsibly, making AI governance both an access problem and a stewardship problem.x

A significant need remains for empirical research that centers district leaders’ governance work as they interpret inequities and ethical risks, design implementation strategies, and coordinate distributed responsibilities in the context of limited external guidance. Although policy reports and emerging reviews identify key concerns and articulate broad principles for responsible AI adoption,xi less is known about how these principles are operationalized in routine district decision-making.

Study Background

Findings and recommendations are informed by qualitative, one-hour interviews with district leaders (N=9) across seven districts in Northern and Central Virginia. Interviews were conducted in the summer of 2025. District leaders included roles such as Assistant Superintendent of Technology and Chief Information and IT Officer. Participating districts are actively engaged in AI implementation. However, the districts were at different phases of AI readiness, with some being early adopters and others just beginning their journey. Furthermore, district leaders represent districts with varying sizes, urbanicity, demographics, and levels of resource access. Further information about participating districts can be found here.

Analysis of interview data was guided by distributed leadership theory,xii which emphasizes how leadership practices are "stretched over" multiple actors, tools, and situations rather than residing in individual leaders.xiii Applied to AI implementation, this theory illuminates how ethical decision-making about AI tools involves distributed practices across organizational levels, from district technology officers to building principals and others, each contributing to the collective leadership of AI governance.

Findings: Key Considerations Impacting Decision-Making for AI Policy

Several themes emerged in our interview data that inform this brief. Those entail navigating compliance and procurement policies, establishing governance structures to inform decision-making, and considering related equity issues.

Finding the balance between compliance and innovation

Compliance, especially the protection of personally identifiable information of students and staff, is central to AI governance and to how district leaders consider AI adoption. However, district leaders noted that they receive conflicting messages and little state or federal guidance about AI governance. Instead, leaders expressed that they are encouraged to rapidly adopt AI tools in their districts without sufficient consideration for compliance structures that typically guide decision-making. For example, a district chief stated:

VDOE has encouraged without real guidance on the actual balance of the security concerns, the ethics concerns, the diversity concerns, the legal concerns. I don't know that the guidance we've gotten from the VDOE and the state to like go out and just use these tools has really taken into consideration the real practical outcomes that then we are on the hook for to make sure that we're really being truly compliant.

Even though Virginia has an AI integration framework for education that outlines basic principles, it is high-level and leaves considerable room for districts to set their own policies. Participants noted concerns that there is not a more coherent state strategy, as this could contribute to inequities between districts in the resources and policies devoted to AI implementation.

Furthermore, leaders expressed operating within a dichotomy of innovation and the rapid growth of AI use, and deep concern about data protection and compliance. For example:

The biggest challenge for me is ensuring that balance between wanting to innovate and use new platforms and ensuring that they are compliant with all of our rules and regulations and federal guidelines for FERPA… a lot of software sales vendors who are happy to come in and give our teachers or anybody it free and then they're stealing our data and using it to train their model. And so we have literally blocked on our network all AI tools unless they are approved.

Leaders noted that they encourage staff to be cautious in their use of AI due to FERPA compliance and other federal regulations, as many technology companies are not transparent about their data policies and ownership. This creates tension: districts want to explore and test new AI tools as they emerge but must operate slowly and cautiously due to potential data security violations.

Because of data compliance concerns, district leaders are revisiting their vetting and procurement policies and developing different approaches to encourage the use and vetting of new technology. For example, one district shared that teachers are encouraged to request new AI tools to enhance instructional innovation. As part of the process, the district developed a rubric to assess the requested AI tool for data security and ethical concerns. The district leader shared their process:

  1. The District’s AI committee vets the teacher’s request for cost, feasibility of use, and any existing alternatives or similar options within the district.
  2. If the request passes the committee, the district-level instructional coordinator drafts a letter that defines why the tool is needed to inform the contract process.
  3. Legal services use language from the instructional coordinator and existing compliance policies to negotiate the contract terms and conditions with the vendor.

As described, numerous departments with specific responsibilities are involved in reviewing teacher requests for AI tools. This layered approach provides multiple touchpoints to distribute responsibilities, ensuring that the tool is compliant, protects student data, and is responsive to teachers’ interests.

Policies Informing AI Policies

There are a number of policies and frameworks that guide district policies and decision-making around AI. These include:

Developing governance structures to increase engagement and distribute leadership responsibilities

As described above, districts use cross-departmental structures to share responsibility for vetting and procuring new AI-enabled tools. Similar structures and processes are also used for stakeholder engagement, as many districts include central office and school-level staff on AI taskforce committees. These committees provide critical input for decision-making. For example, one of the districts established a committee of 10 staff members from different departments and roles to annually review generative AI tools, using a ranking system designed by the Chief IT Officer to assess the tool's safety. The Chief shared:

With generative AI, we have a team that's focused on a yearly review of those tools. They're coming so fast and even embedded into other tools. And so, we really want to have visibility into that. We're formalizing that review process now and we'll be using some rankings for things that are low risk, medium risk, high risk and really focusing on the high-risk applications that potentially could have a negative impact to a student if they have the wrong type of chat.

Through this established structure, the Chief can control which tools enter the classroom and assess risk levels using defined parameters set by other trusted leaders. Importantly, this process engages diverse stakeholders in the review process to gather their feedback and include them in decision-making.

The development of an AI-related steering committee is typical across districts; however, the structure, engagement, and purpose differed by district. In some districts, the committee creates protocols and guardrails to support all teachers in exploring AI tools for instructional purposes. Based on teacher feedback, the committee decides which tools to adopt. For one mid-size district, a committee of 40 educators formed to review and test the top 10 AI tools the district considered adopting, and a parent advisory committee on educational technology formed to gather additional stakeholder feedback. Other districts are not as exploratory with educators and instead engage committees in compliance checks.

These committees often begin small and grow over time to become more inclusive and representative of more departments and schools. One mid-size district described forming an AI steering committee that began as a group of five people and grew to be cross-departmental, including nearly all central office departments, and representation from school administrators through cabinet-level members. The office of DEI was actively involved in building the committee, which supported a focus on equity in discussions of AI adoption decisions. For smaller districts, it is easier to be inclusive of school staff, with one district reporting a steering committee comprised of staff from every school in the district, as well as cross-departmental representation from central office. Having these committee structures in place enables deeper collaboration and greater nuance in decision-making around AI adoption.

Differences in conceptualizing and applying equity principles

Districts hold different ideas about how equity is conceptualized and applied in AI adoption and governance. Some districts focus more on equity in access, ensuring that all students have opportunities to build critical AI literacy skills, with a strong commitment not to exacerbate inequities based on resource availability. For example:

This is a massive technology. Powerful. It needs to be done the right way, right? But that creates an equity gap that we didn't have before. Kids who are going home and using tools like this... And if we're not allowing some tool for all of our kids, then what are we doing?

Many district leaders feel a deep responsibility to ensure that all students have access to AI tools to narrow the opportunity gap between those with access and those without.

These districts also share an equity concern about educators' capacity and knowledge regarding AI tools. District leaders discussed tension among staff, as some educators fully embrace AI and incorporate it into their practice, while others are not in favor of its use. The lack of cohesion in use and capacity among some staff can lead to inequities in the classroom when access to tools and knowledge is unevenly distributed. These emerging divisions create a sense of urgency to train staff in AI use so that some are not left behind, and others do not become frustrated by a lack of progress or by limitations to their practice. Districts use professional development and encourage educators to test AI tools to address this issue.

Other districts discussed equity in terms of equitable systems to collect, protect, and analyze data for equitable decision-making and are heavily invested in using AI tools to facilitate this. These districts prioritize using AI to work towards more equitable learning conditions through equity-informed decision-making by investing early in establishing data systems that support advanced AI-enabled analytics. For example, one district leader shared:

Ultimately, all of that will enable the use of AI tools and will put experts who can then guide the organization in the use of those tools and effectively will ensure that we have good quality data for decision making...We want to make sure that as we start to make decisions down to like the individual student level or looking at barriers for types of students towards the things that might be getting in their way...

Examples of the types of data analysis and equity issues that could be uncovered include access to advanced coursework and leveraging AI to identify issues in earlier grades that could negatively impact a student in later grades, to help access educational opportunities. Another district leader from the same district shared a similar idea of equity and AI, saying, “AI can either help if you are not intentional about considering all students [in your data], or it can help expose [inequities] if you are intentional.”

“Kids Were Cheating 10 Years Ago, 20 Years Ago”

Ethical concerns regarding plagiarism and AI policies remain a point of contention in AI adoption and governance. District leaders reflected on the problems of practice that could unintentionally contribute to cheating when using AI. One district leader shared:

Kids were cheating 10 years ago, 20 years ago. They're going to cheat tomorrow. But I think that makes us think. It's making us have good conversations about what our instructional goals are and what our assessments are and how does that align with our instructional goals.

This quote reflects a commonly held view among district leaders. Districts should adapt their approach to student assessments to ensure that what students produce aligns with instructional goals, regardless of whether AI is used. This could mean that students demonstrate their knowledge through new mechanisms that do not lend themselves to copying and pasting information or AI-generated material. This approach to plagiarism and AI places responsibility on educators, rather than assuming students will be bad actors. When responsibility falls on students, it can lead to plagiarism policies that unintentionally contribute to inequities by misidentifying student work as AI-generated.

Recommendations

Based on findings from qualitative interviews, districts can adopt a range of practices and policies to address issues in AI governance and equity.

Create a tiered access model for AI use in the classroom. The model could include specific requirements for acceptable use and permission to access the AI tool, depending on the student's age and purpose of use, as well as on whether the use is by an educator or a student. This addresses concerns related to cohesion and unequal adoption. The tiered access model could also help inform vetting standards with industry partners and identify the number of users per tool to improve efficiency. It also provides educators with clear parameters for classroom use.

Invest in AI literacy for both staff and students. In many ways, district leadership emphasized the importance of AI literacy as a strategy for advancing AI policy implementation. This does not mean mere access to tools but rather access to meaningful learning and engagement so that AI tools are used equitably and ethically.

Establish professional use policy. Districts have focused mostly on AI policy for the classroom since protecting students is of the utmost importance. In doing so, a gap emerged in AI policy for professional use. While district leaders recognize the efficiencies of AI for administrative work, little guidance has been rolled out on acceptable use in this space.

Several recommendations also emerged for state education leaders.

Centralized tool vetting would improve efficiency and equity. District leaders noted a lack of guidance on tool vetting and a need for more centralized state-level governance to support it. This was raised as an equity issue, since some districts might adopt a tool that is not allowed in other districts, causing inequities in student and teacher access.

Establish state standards and assessment tools for AI. These should measure the growth of AI literacy and use among educators and students across the state. This is a critical step to monitor potential gaps in access and opportunities for under-resourced communities, as well as to protect against the misuse of AI in education.

Conclusion

Even though the state issued more guidance and legislation on AI use in education since we collected interview data, one thing is clear. District leaders are navigating complex issues, including data security, curriculum alignment, staff capacity building, and ethical concerns. They are responding by revamping their procurement policies, forming new governance structures, and conceptualizing what equity means in AI for education. As AI in education continues to unfold, it is critical that the structures and policies developed are nimble and responsive to the ever-changing landscape of both policy and technology. This requires additional engagement and collaboration between the state and districts, as well as stronger state guidance that centers equity for all.

This brief is part of a series in partnership between George Mason University’s AI-LIFT (Learning, Innovation, and Future Technology) and EdPolicyForward: The Center for Education Policy at George Mason University’s College of Education and Human Development. The authors wish to express their gratitude to the partnering school districts for their collaboration.

  1. Lake, R. & Makori, A. (June 2020). The digital divide among students during COVID-19: Who has access? Who doesn’t? The Center on Reinventing Public Education. Retrieved from: https://crpe.org/the-digital-divide-among-students-during-covid-19-who-has-access-who-doesnt/
  2. U.S. Department of Education, Office of Educational Technology. (2023). Artificial intelligence and the future of teaching and learning: Insights and recommendations. Retrieved from: https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf
  3. Diliberti, M. K., Lake, R. J., & Weiner, S. R. (2025). More districts are training teachers on artificial intelligence: Findings from the American School District Panel. RAND Corporation. RAND Corporation. (2025). AI use in schools is quickly increasing but guidance lags behind: Findings from the American School District Panel.
  4. U.S. Department of Education, Office of Educational Technology. (2023). Artificial intelligence and the future of teaching and learning: Insights and recommendations. Retrieved from: https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf
  5. UNESCO. (2023). Guidance for generative AI in education and research.
  6. Ho, C. S. M. (2025). Principals’ ethical leadership in the AI era: A narrative literature review of emerging challenges, strategies, and outcomes. Computers & Education, 243, 105517. https://doi.org/10.1016/j.compedu.2025.105517
    National Institute of Standards and Technology. (2023). Artificial intelligence risk management framework (AI RMF 1.0) (NIST AI 100-1).
  7. Cheah, Y. H., Lu, J., & Kim, J. (2025). Integrating generative artificial intelligence in K–12 education: Examining teachers’ preparedness, practices, and barriers. Computers and Education: Artificial Intelligence, 8, 100363. https://doi.org/10.1016/j.caeai.2025.100363
  8. Lake, R. & Makori, A. (June 2020). The digital divide among students during COVID-19: Who has access? Who doesn’t? The Center on Reinventing Public Education. Retrieved from: https://crpe.org/the-digital-divide-among-students-during-covid-19-who-has-access-who-doesnt/
  9. Pham, H., Kohli, T., Olick Llano, E., Nokuri, I., & Weinstock, A. (2024, June 29). How will AI impact racial disparities in education? Stanford Center for Racial Justice. Retrieved from: https://law.stanford.edu/2024/06/29/how-will-ai-impact-racial-disparities-in-education/
  10. ibid
  11. Ho, C. S. M. (2025). Principals’ ethical leadership in the AI era: A narrative literature review of emerging challenges, strategies, and outcomes. Computers & Education, 243, 105517. https://doi.org/10.1016/j.compedu.2025.105517 UNESCO. (2023). Guidance for generative AI in education and research.
  12. Spillane, J. P., Halverson, R., & Diamond, J. B. (2004). Towards a theory of leadership practice: A distributed perspective. Journal of Curriculum Studies, 36(1), 3–34.
  13. Ho, J., Hung, D., Chua, P. H., & Munir, N. B. (2022). Integrating distributed with ecological leadership: Through the lens of activity theory. Educational Management Administration & Leadership, 50(4), 647–665.