Online learning provides flexibility and accessibility, but creating a sense of belonging, connectedness, and fairness continues to be a challenge. This study investigates the potential of an AI-powered social agent, Social Agent Mediated Interactions (SAMI), to address challenges in fostering inclusivity and equitable access in online learning. It examines how learners’ motivation to engage with AI tools is shaped by their need to belong (NTB), self-efficacy (SE), and personality traits. By analyzing the interactions between these factors and AI adoption, we assess how SAMI can be adapted to diverse learner profiles, promoting more inclusive and supportive online learning experiences.

This study explores the potential of SAMI, a novel AI-based intervention, to enhance students’ sense of belonging and community in online learning environments. SAMI facilitates social connections by recommending peers based on information from learners' self-introductions, which are created at the beginning of the semester and include details such as hobbies and location. Our research examines SAMI’s deployment in the online Knowledge-Based Artificial Intelligence (KBAI) course at the Georgia Institute of Technology across multiple semesters.

We investigate whether SAMI can foster a stronger sense of belonging (SOB) and engagement among learners, as well as how demographic and personality factors influence their adoption and experience. Specifically, we examine how learners’ motivation to engage with AI tools is shaped by their NTB, SE, and personality traits. By analyzing the interaction between these factors and AI adoption, we aim to determine how SAMI can be optimized for diverse learner profiles, ultimately creating a more inclusive and supportive online learning environment.

We anticipate a life cycle where the NTB serves as the motivation, a sense of belonging as the outcome, and SE as an influencing factor. We expect the motivation to use social agents and seek belonging within course settings can arise from factors beyond NTB and SE, potentially driven by personality traits. That leads us to our several key dimensions of analysis:
- SOB: Does the implementation of SAMI promote a stronger sense of belonging among learners enrolled in the online KBAI course?
- NTB: Do learners who score higher on NTB choose to use SAMI? How does NTB influence a learner’s intention to engage with SAMI?
- SE: Are there measurable differences in self-efficacy between students who utilize SAMI and those who do not?
- Big Five personality traits: How does personality impact learner’s motivation to engage with AI-tools?

We also consider learner demographics in the context of:
- Equity in Adoption: Is there a disparity in learner adoption and motivations of SAMI based on their demographic characteristics?

By examining these factors, we aim to understand SAMI’s role in fostering community, supporting learning, and potentially mitigating feelings of isolation in online education. Our research indicates there is no bias in learners’ adoption of SAMI (decision to opt in) across various demographics, promoting an equitable environment where all learners feel comfortable using the tool. Our findings suggest that SAMI has the potential to enhance learners’ sense of belonging and community in online environments. However, we observe no significant differences in the need to belong or self-efficacy between users of SAMI and non-SAMI adopters.

Keywords: AI social agents, Community of Inquiry, social presence, social interaction, learning efficiency, learning engagement, online learning, personality traits.