AI as an Extra Set of Eyes: Using AI Observation and Analytics To Support Student Success

When used responsibly, these capabilities provide educators with another source of information; however, they should never replace professional judgment. “In education, trust is foundational,” says Matt Jubelirer, general manager for education marketing at Microsoft. “Schools should never have to choose between innovation and responsibility.” 

That balance is especially important when schools use AI for student monitoring or analytics. Jubelirer says Microsoft’s goal is “to help teachers identify opportunities for student support, personalize learning experiences and reduce administrative burden while ensuring that people remain at the center of decision-making.” 

Balancing AI Use, Privacy and Trust 

Building trust around the use of AI requires clear communication. Educators, students and families should know how AI is used, what information it collects, the safeguards in place and when human review is required. Microsoft’s approach to AI in education is guided by three principles: human-centered design, transparency and educator control, and privacy and security by default. 

Jubelirer says that Microsoft implements safeguards such as “content filtering, dynamic blocklists for inappropriate prompts and additional protections designed to help ensure AI interactions remain safe, age-appropriate and relevant for educational settings.”  Within Teams for Education, features such as Expected AI Use and Restricted Mode allow educators to communicate assignment-level expectations and lets both administrators and educators manage student access to AI capabilities, he says. 

However, technology safeguards are only one part of responsible implementation.

Sarah Thomas, founder of EduMatch, advises, “Before adopting any AI-driven monitoring or analytics tool, schools should know what data they collect and who will be able to see it.” Personally identifiable information should be taken seriously, and districts should have a data privacy agreement with any tool that could potentially touch student and staff data.  

DIVE DEEPER: Data governance helps K–12 schools safeguard student data.

Schools must look beyond a vendor’s general privacy statement and determine what data enters the system, who has access, whether it is stored or shared, and what happens to it when the district stops using the product. 

Jubelirer emphasizes three considerations for institutions thinking about adopting or implementing AI: governance of use, governance of data and governance of risk and oversight. He adds, “Effective frameworks provide practical guidance for use, ensure responsible data management and privacy, and enable ongoing risk oversight as technology and institutional needs continue to evolve.” 

These considerations should be addressed before implementation. Schools must define appropriate uses, establish accountability, create reporting procedures and regularly review whether the technology continues to serve its educational purpose.

Jubelirer emphasizes the need for “clear expectations, transparency about how AI systems are used, professional learning for educators and policies that prioritize responsible use.” Educators, students and families should also have opportunities to ask questions and contribute to decisions about how these systems are used.

AI Should Inform Decisions, Not Make Them

An algorithm may identify patterns, but it cannot understand the complete picture of a student’s life and learning experience. It also cannot fully account for family circumstances, social-emotional factors, classroom relationships, accessibility needs or the many experiences that influence learning. Applying context, empathy and professional expertise remains the educator’s responsibility, Jubelirer says: “AI can be a valuable assistant, but it should not be the sole basis for consequential decisions about students, learning outcomes or well-being.” 

For example, a learning analytics platform may identify that a student’s assignment completion, class participation and assessment performance have declined over a few weeks. AI-generated information should prompt questions, not be treated as a conclusion. Rather than automatically labeling the student as “at risk,” the system should alert the teacher, who can evaluate the information for accuracy, consider possible reasons for the change in behavior, apply relevant context and, most important, engage in a conversation with the student. 

Christopher Hoang, director of educational technology and innovation for the Los Angeles County Office of Education, cautions, “If we don't focus on understanding students and their needs, and instead keep focusing on data, we neglect the human aspect of education. AI-powered tools can enhance the teaching experience by focusing on the student experience.”

WATCH: These K–12 districts are building AI and UDL into every classroom.

Policies and Professional Learning

Districts must clearly communicate guidelines for AI implementation. Schools need professional learning that helps educators understand what AI systems do, their limitations and how to recognize potentially biased or misleading conclusions. 

Because AI continues to change rapidly, this learning cannot be limited to a one-time product demonstration. Clear policies are essential, but schools must convert those policies into practical classroom guidance for daily instruction and decision-making. “The most successful implementations are those where technology, training, transparency and accountability work together,” says Jubelirer.

This includes defining appropriate uses, establishing educator oversight, creating procedures for addressing concerns and communicating expectations to students and families.

Keeping Relationships at the Center

Nneka McGee, founder and strategic adviser for Muon Global, says, “Humans should always determine the parameters under which the tools will be used and the outcomes they hope to achieve with their use. Human-led evaluation is necessary to determine if goals are met with AI-powered tools. There are programs such as outcomes-based contracting that support these types of initiatives.”

For schools beginning this journey, the Microsoft Education AI Toolkit offers guidance for developing AI strategies, governance approaches and implementation plans aligned with educational goals. As AI capabilities evolve, educators should rethink how technology can support teaching, rather than replace it. 

When implemented thoughtfully, AI analytics can help identify students who need assistance, enable earlier intervention and reduce administrative burdens. However, the value of these tools ultimately depends on whether they strengthen human connections.

“AI can help us find patterns and transform our teaching, but as the great Rita Pierson said, ‘every child needs a champion,’” Thomas added, speaking of the longtime educator and teacher advocate. “Relationships are the core of learning. AI is a tool that can help us serve our students better.”

Hoang reinforces that point: “The fact is, AI, like any tool, is only as good as the user. AI won't replace great teaching; it will empower those who are great teachers to be even greater teachers.”

The goal is not for AI to watch students. It is to help educators notice what matters sooner, ask better questions and make informed decisions that support student success. When AI serves as an additional set of eyes rather than a substitute for human judgment, it can amplify the expertise, compassion and relationships that remain at the heart of great teaching.