CARLS Learning Systems
Human-Centered AI-Assisted Learning
About CARLS Learning Systems
CARLS Learning Systems develops human-centered approaches to AI-assisted learning through the Cognitive Adaptive Regulation and Learning System (CARLS)—a research-driven framework designed to help learners regulate, reflect, engage thoughtfully with AI assistance, develop AI literacy, and build greater independence over time. The framework keeps educators at the center of instructional governance while using governed learning information to help teachers understand how students are progressing and where individual support may be needed. This can help educators focus one-to-one attention where it is most valuable and make informed adjustments to instruction, while maintaining teacher control over the context and boundaries within which AI assistance occurs.
CARLS Learning Systems was founded by Selena Singh, a Human-Computer Interaction & AI Governance Research Scientist. Her work focuses on the design and governance of human-centered AI systems that support rather than replace human judgment and agency. She developed the CARLS framework as part of this research, which has been accepted for presentation at IEEE AIxHEART 2026. CARLS is now progressing toward prototype development and empirical evaluation.
Our Research and Development Focus
Regulation & Reflection
Guided regulation supports mental and emotional readiness for learning, helping learners settle their attention before engaging with academic tasks. Students then reflect on their understanding and use CARLS’ structured question-and-answer process to formulate meaningful questions before seeking AI assistance—supporting thoughtful engagement and greater confidence in their own learning process.
Governed Assistance & Teacher Insight
Educators establish the boundaries for AI assistance while CARLS captures governed learning indicators that can help teachers identify where individual students may need additional support. Learners practice questioning, evaluating, and applying AI-supported information to build AI literacy rather than passive reliance.
Recalibration & Growing Independence
CARLS helps educators understand how learners are progressing and where individual support may still be needed. As learners build understanding and confidence, support can be recalibrated accordingly. Success means learners become increasingly able to reason through problems, explain and apply what they have learned, and complete work without needing AI assistance.