Spotlighting Emergent AI-enabled Literacy, Math, and Belonging Practices from the School Teams AI Collaborative

We have reached the midpoint of the School Teams AI Collaborative, a joint effort between Leading Educators and The Learning Accelerator to position real educators to lead in the AI era. This offers us a powerful moment to reflect, learn, and set the stage for the next phase of AI-powered instructional innovation.

While the field at large is theorizing about the potential role that AI could play in education, the 19 school teams in the Collaborative are learning by experimenting with it within instruction. It’s not a mere add-on or trend: it’s a new lever for working toward what we have always known excellent teaching should be.

At our virtual mid-year convening, three key themes emerged that underscore what the school teams are learning about using AI: centering human experience, enhancing critical thinking and engagement, and using a process of testing, learning, and adapting. 

• Centering human experience emphasizes AI’s role in amplifying meaningful teacher-student interactions and creating personalized learning pathways.
• Enhancing critical thinking and engagement focuses on using AI as a tool to deepen inquiry, promote metacognition, and encourage active problem-solving rather than using AI to do the heavy lifting for students.
• Finally, the test-learn-adapt process highlights the importance of an iterative approach—using data to refine strategies, ensure effectiveness, and scale impactful AI practices.

The following three educator spotlights highlight powerful examples of how these themes are coming to life in concrete AI-driven teaching practices. Educators are pushing boundaries, centering student needs, and refining their strategies through real-world implementation.

Spotlight 1: Centering Human Experience

Troy Valentine: Denver Online – Middle and high school Civics and 8th grade US history teacher

Denver Public Schools’s Denver Online serves as a safe haven for students who have faced challenging circumstances, learning differences, or simply need a more flexible educational model. With student agency and choice at its core, the school provides an adaptive, blended learning structure that meets students where they are.

⚡ About the AI application: At Denver Online, educator Troy Valentine has explored AI’s potential to create deeper student connections in virtual learning. While chatbots are often used for instructional support, Troy discovered that they also helped students feel safe expressing themselves in unexpected ways.

One of his students, who has struggled with social engagement and rarely speaks up in class, used the chatbot not just for academic support but as a space to share personal reflections. 

While reviewing chatbot transcripts—a feature that allows teachers to monitor student interactions—Troy noticed the student had mentioned a personal goal: they wanted to step outside that day, a small but significant step for someone facing deep personal challenges. This was something the student likely would not have shared with peers or in class, yet through the chatbot, they felt comfortable expressing it. This moment opened a door for Troy to connect with the student more directly, strengthening their relationship and providing needed support.

• Beyond relationship-building, Troy found that chatbots allowed students to explore topics they were curious about—asking questions beyond the assignment prompts, diving deeper into civic issues, and shaping their own learning journey.
• The ability to track these interactions, with student and parent/guardian consent, helped Troy understand what genuinely interested his students, guiding future instruction in a more personalized and meaningful way.

💡 The lesson: Troy’s experience underscores that AI may serve as a touchpoint for students, but it’s the educator who ultimately fosters the relationships and engagement that make learning transformative. By using AI strategically, Troy has created a more supportive, student-centered environment that prioritizes human relationships, connection, curiosity, and emotional well-being.

Spotlight 2: Enhancing Critical Thinking and Engagement

Tara Correa: Eliot K-8 Innovation School – 6th grade literacy

At Boston Public Schools’ Eliot K-8 Innovation School, ELA teachers are using AI to provide targeted, growth-oriented feedback on writing, allowing students to engage in metacognitive reflection.

⚡ About the AI application: AI-generated feedback is just one step in a holistic instructional practice. Students identified their strengths and areas for improvement in the feedback, using graphic organizers to create writing plans that prioritized their most high-impact revisions. 

• One key benefit of this approach is its impact on metacognitive skill-building—students are learning to analyze their work beyond surface-level corrections, strengthening arguments, and refining ideas.
• Teachers can also use AI-generated feedback trends to guide small group instruction, ensuring students receive tailored support to push their thinking.

In the future, the teachers are planning an additional meta-reflection opportunity for students to analyze their feedback over time, looking for how the feedback has shifted or what areas stayed the same. This reflection allows them to make informed goals for themselves as writers in future assignments. Then, AI feedback could potentially be generated specific to their own writing goals.

💡 The lesson: Eliot K-8’s approach exemplifies keeping humans in the loop. Educators use AI to enhance and strategically direct, not replace, instructional guidance—helping students reflect, adapt, prioritize, and develop skills that will serve them beyond the classroom. 

Spotlight 3: Test, Learn, Adapt

Sam Rhodes: Georgia Southern University – Education researcher and professor
Varuni Tiruchelvam: CIOB Citywide District – Instructional coach

Our third spotlight takes us beyond the Collaborative to explore AI’s role in math instruction.

Our guests, Sam Rhodes, an education researcher at Georgia Southern University, and Varuni Tiruchelvam, an instructional coach in NYC’s Consortium, Internationals, and NYC Outward Bound Schools (CIOB) Citywide District, partnered with math teachers Janice Trinidad and Jenny Zheng from Cedars NextGen High School in a research sprint facilitated by Playlab, a nonprofit that enables educators and impact organizations to build AI-powered tools and experiences. 

⚡ About the AI application: Their goal was to foster metacognition by using an AI assistant to ask students guiding questions rather than providing direct answers, encouraging deeper thinking through a structured four-phase approach inspired by mathematician George Pólya: understanding the problem, planning a solution, executing the plan, and reflecting on the process. 

• To test the tool, the team surveyed students about how helpful they found it and how it made them think.
• They also examined student outcome data, comparing the number of students who got the correct answer on the problem with the most AI tool use cases (15/16) to a comparable problem without the tool (9/16).
• Finally they analyzed the activity logs of the students engaging with the tool to assess the quality of the student prompts and interactions and where the AI was effective and ineffective. 

Early findings showed that students who used the tool demonstrated stronger conceptual understanding and problem-solving accuracy compared to those who did not. However, challenges arose in ensuring students fully understood how to engage with the AI effectively, leading to refinements in both the tool and instructional strategies. Moving forward, the team aims to refine the AI’s questioning strategies to gradually shift responsibility to students, fostering independent thinking.

💡 The lesson: Having multiple data points across qualitative and quantitative measures gave the team a full picture of what was working and what needed to be adjusted, and a way to assess the effectiveness of the adjustments they were making to the tool and the instructional practices around the tool. Their work highlights the importance of iterative learning—continuously testing, refining, and adapting AI tools to enhance, rather than replace, student engagement in math learning.

Deprivatized Practice Supports Common Learning

While no two schools are the same, the exchange of ideas like these can inspire new ways of thinking that cross contexts.

Insights from other educators—whether working in similar or vastly different settings—serve as catalysts for experimentation and growth. While a strategy that works in one classroom may not be directly replicable elsewhere, it can inspire adaptations that lead to groundbreaking new approaches.

Diving Deeper

After hearing these inspiring spotlights at the mid-year convening, Collaborative participants engaged in deeper discussions around the core themes:

1. Centering human experience: How can AI elevate the human aspects of learning and teaching?
2. Enhancing critical thinking and engagement: How can educators and students use AI to deepen student inquiry and problem-solving rather than using it to do the thinking?
3. Test, learn, adapt: How can we apply a data-informed, experimental approach to our AI implementations to ensure that we are driving toward meaningful student outcomes?

These discussions allowed participants to process ideas, share open-ended questions, and explore practical applications of the insights gained. Conversations remained personal and focused, fostering deeper connections and collaboration. The key takeaway? AI is not a magic solution—it’s a tool that, when intentionally designed and thoughtfully implemented, can enhance learning in powerful ways.

Likewise, teaching and learning are continuous processes of iteration, reflection, and adaptation. By sharing what works, questioning assumptions, and remaining open to experimentation, we ensure that AI serves as a meaningful force for student-centered learning.

Let’s keep learning, sharing, and pushing forward—together.