A Public-Source Comparative Case Study of the DIGS Framework for AI-Enabled Academic Advising Platform Selection

Abstract

This public-source comparative case study examines whether the Digital Infrastructure Governance and Selection (DIGS) framework can support defensible decisions about AI-enabled academic advising platforms in higher education. Developed by Fatema Akter, DIGS combines noncompensatory assurance requirements with six lifecycle stage gates. The framework responds to three weaknesses in conventional procurement: aggregate scores can conceal failed safeguards, vendor assertions are often treated as verified evidence, and purchase decisions are separated from implementation outcomes. We applied seven decision domains to publicly available documentation: mission and service fit; architecture and interoperability; data governance and privacy; cybersecurity and resilience; accessibility and equity; lifecycle economics and viability; and implementation and benefit realization. Public evidence proved too limited for a complete governance audit. Advisor.AI was the only platform in the reviewed material with a dedicated product page, while disclosures about large language model providers, privacy compliance, cybersecurity certification, and accessibility were absent. The economic evidence was more informative. The sources attributed 70-85% of total AI cost to the operating environment and indicated that recurring run costs can exceed cumulative build costs within 12 to 24 months. A simulated fairness audit using the UCI Student Performance Dataset also produced differences in predicted positive rates by gender and parental education. These results show both the value and the limit of public-source evaluation: DIGS makes missing evidence visible, but institutions still need contractual disclosure, independent assurance, and controlled pilots before approving a platform.