Research Summary: Accessible, but Not Adopted: Increasing LLM Adoption among First-generation, Low-income (FGLI) College Students beyond Expanding Access

Research indicates that while large language models (LLMs) are increasingly available to underserved communities, including first-generation, low-income (FGLI) college students, mere access does not guarantee meaningful adoption. The study highlights a critical 'access-adoption gap,' suggesting that factors beyond availability influence whether users are willing to adopt LLMs and deeply integrate diverse LLM tools into their practices. This gap hinders the full realization of LLMs' social potential.

Why it matters

This analysis is critical for stakeholders investing in technology to empower diverse populations, as it underscores that successful initiatives require more than just providing access. Understanding and addressing the 'access-adoption gap' is essential for maximizing the return on investment in new technologies and ensuring their social benefits are fully achieved across various societal segments.

Key insights

• LLMs are being positioned as tools to empower underserved communities.
• Significant efforts are underway to expand access to LLMs.
• Access to LLMs alone does not equate to meaningful adoption by users.
• User willingness is a key factor in LLM adoption, even if systems are accessible.
• The heterogeneity of LLM tools means adoption can be deepened beyond superficial use.
• A critical 'access-adoption gap' exists, preventing the full social potential of LLMs from being realized.
• The findings are drawn from 61 interviews, including 15 long-form semi-structured interviews with FGLI students.

Source

arXiv — Computers and Society — https://arxiv.org/abs/2609.36129