Schools have no shortage of data. Learning management systems record access patterns, assignment submissions, quiz attempts, and student activity. Digital tools capture usage trends, while help desks collect reports about recurring technical problems. Surveys, attendance records, and assessment platforms add even more information.
Much of this information, which educators often describe as dark or underused data, is unused or only partially examined, and may hold important clues about teaching, learning, and student support. However, the data itself has no value until it is engaged within an instructional or leadership context.
Generative AI has increased interest in making better use of these large collections of information. School leaders, instructional coaches, and technology facilitators are beginning to ask whether AI can help process extant data more quickly and identify patterns that would otherwise remain buried. That interest is understandable as AI can review large volumes of text, sort information into categories, summarize findings, and highlight trends that deserve closer attention.
Article continues below
The real opportunity is not simply faster analysis, but smarter analysis guided by professional judgment. This is where educators need to be careful. AI can help schools become more data informed, but it can also tempt them into becoming overly dependent on automated conclusions.
Education has always required interpretation, context, and human understanding. That has not changed, and in fact, human judgment might be more important today than before the expansion of generative AI.
Contextualizing Data
Numbers and patterns matter, but need context. A dashboard cannot understand a student’s home responsibilities. A large language model cannot fully grasp classroom culture. A summary report cannot replace the experience of a teacher or coach who knows the learners behind the data.
A data-informed approach is more useful than a purely data-driven one. Data-driven decision-making often suggests that the numbers point directly to the answer, while data-informed decision-making treats data as one important source of evidence among several. Educators still need to ask what the information means, what it does not capture, and how local context should shape the response. That distinction matters even more now that generative AI can synthesize so much information so quickly.
AI is especially useful during the early stages of inquiry as it can help instructional leaders look across multiple sources of existing data and spot areas worth investigating. A coach might use AI to review comments from teacher surveys and identify recurring concerns about student engagement. A technology facilitator might use it to sort support tickets into categories and discover that a particular platform is creating repeated confusion. A school leader might compare patterns in assignment completion, attendance, and platform usage to identify students who may need additional support.
These are meaningful uses of AI because each helps to identify patterns at scale. It does not remove the need for human interpretation.
A spike in late-night platform activity might look like procrastination, yet it might also reflect sports schedules, work commitments, family care responsibilities, or uneven internet access at home. Low use of a digital tool might suggest weak implementation, or that the tool possibly does not align with the lesson goals, or that a teacher found a better non-digital strategy.
Limited revision activity in student writing might signal low engagement. It might also mean students are drafting in notebooks, using a different platform, or receiving verbal feedback during class.
The pattern is only the starting point. The judgment comes next.
Asking the Right Questions
Instructional coaches and technology facilitators are especially well-positioned to support leaders in this work as they often sit at the intersection of classroom practice, digital systems, and professional learning. They can help schools move beyond surface-level dashboards and ask better questions about the data they already have. They can also help teams resist the urge to treat AI output as final truth.
One of the most important roles these educators can play is framing the right inquiry question before any analysis begins. Instead of asking AI to find everything interesting in a large data set, schools should begin with questions tied to teaching and learning, such as:
- Are students using feedback to improve their work?
- Are teachers using digital tools in ways that support differentiation?
- Are families receiving communication in forms they can access and understand?
- Are accessibility features reaching the students who need it the most?
Clear questions produce more useful analysis, and human-centered ones produce more ethical analysis. Generative AI should serve as an assistant, not a decision-maker. It can surface possibilities, summarize evidence, and identify trends that deserve discussion. It cannot determine the right instructional response on its own. That work still belongs to educators who understand the curriculum, the learners, the community, and the goals of the school.
This is where leadership matters. School and district leaders need to create expectations for responsible AI use in data analysis. Staff should know what kinds of data are appropriate to analyze with AI tools, what privacy protections must be in place, and how human review will remain part of the process. Teams should be encouraged to validate AI-generated findings against direct observation, classroom artifacts, and professional conversation. Trustworthy use of AI depends on thoughtful structures, not just powerful tools.
The schools that benefit most from AI will not be the ones that automate the most decisions, it will be the ones that use it to ask sharper questions, uncover overlooked signals, and support better conversations about teaching and learning. Those schools will use extant data more effectively without forgetting that education is a human enterprise.
Human judgment is not a barrier to innovation, but what makes innovation worth using in the first place.