Why Institutions Must Fix Their Data Before AI Defines Them

 By: Scott Cheney, CEO, Credential Engine

One of the biggest potential challenges facing higher education is being ignored.

Amid the myriad of external crises impacting higher education — from the unknown impacts of changes at the Department of Education to significant declines in enrollment in the years to come — there’s one looming crisis that has received little to no attention.

And it’s one that higher education leaders and policymakers can proactively solve: The crisis of bad data distorting the value of degrees and credentials in an AI-based information landscape.

While well-intentioned, the primary conversation about AI’s greatest impact on higher education remains misguided, with headlines, debates, and finger-wagging overwhelmingly focused on how and whether students will use the technology to cheat. This focus on AI-powered academic dishonesty distracts from the much bigger issue of how AI can shape the way students and families evaluate higher education itself.

Millennials and younger generations already use AI tools at higher rates for everything from basic queries and task management to more complex research — and projections show that generative AI platforms are set to become the primary search tool for more than 90 million Americans by 2027. This means that in the coming years an increasing number of prospective students and their parents will turn to AI — either knowingly or not — to help guide their college, credential, and career pathway decisionmaking.

This has the potential to impact institutions much more deeply than the short-term effects of students using AI to generate essays or come up with answers to staid classroom exercises. Why? Because the outputs of generative AI are only as good as the data available to these tools. While improving, researchers have found that as many as 46% of all responses from commonly-used large language models contain factual inaccuracies. 

How much do you trust an AI bot to communicate the value of an institution’s degrees or credentials? 

Only 25% of Americans believe that a four-year college degree is extremely or very important for obtaining a well-paying job in today’s economy. With so many people already questioning the ROI of higher education, these doubts will be reinforced rather than dispelled if AI tools work with incomplete or inaccurate data.

If accurate and trustworthy data on the value and qualifications an institution’s credentials offer are tucked away in PDFs, spreadsheets, or files hosted on a buried institutional webpage, then AI search tools will quite likely either ignore these completely or misrepresent them. Either way, this leaves prospective students and the general public with incomplete, outdated, or flat-out wrong information about what these programs offer, where they lead, and the skills obtained.

But there’s an easy path for institutions to address these issues head-on. 

Publish Institutional Data in Verified, Open, and Structured Formats

To ensure AI platforms surface accurate and relevant information, institutions themselves should publish their own data in a verified structured format that communicates a wide range of education and workforce information, like the Credential Transparency Description Language (CTDL). This allows AI-driven search tools to recognize and connect degree programs with the competencies and fruitful career pathways they support — making it easier for prospective students to make informed decisions. 

While translating institutional data into an open data format isn’t an overnight task, it doesn’t have to be overwhelming. You can start small by publishing data on ten key credentials or degrees to demonstrate the impact. From there, it’s possible to scale efforts across schools, departments, or entire systems — just as the New Jersey Council of County Colleges is doing by structuring data for more than 1,000 programs, courses, and competencies across the state’s 18 community colleges.

Advocate for Policies That Standardize Credential Data and Transparency

To fully realize the potential of structured, open data, institutions should also advocate for policies that make standardized credential transparency the norm rather than the exception. A patchwork of different credential definitions, inconsistent data formats, and varying levels of detail create unnecessary confusion for students, employers, and AI-powered search tools alike.

Institutional leaders can push for state and federal policies that encourage or require institutions to publish credential data in common, machine-readable formats like CTDL. They can also work with accreditation bodies and employer groups to establish shared standards for describing credentials, competencies, and learning outcomes. By shaping these policies early, institutions can help create a higher education system where AI tools surface accurate and student-friendly information rather than reinforce existing skepticism about the value of a degree.

The institutions that move swiftly to make their data open, structured, and aligned with workforce needs will be best positioned to compete for a shrinking pool of prospective students. As AI continues to reshape how people seek and evaluate information, higher education is in a critical window of opportunity to ensure programs are accurately represented.

Rather than focusing narrowly on AI’s short-term classroom challenges, institutions must take charge of the data used in shaping how AI-powered tools communicate the value of their credentials — before those narratives are written for them.

Scott Cheney is the Chief Executive Officer of Credential Engine, a non-profit on a mission to map the credentials and skills landscape with clear information. Prior to Credential Engine, Cheney served as the Policy Director for Workforce, Economic Development and Pensions for Senator Patty Murray and the Senate Health, Education, Labor and Pensions (HELP) Committee.

All Photos Credited To Unsplash