Editorial
White Paper
Why Enterprise AI Fails at Scale
“Student success” has become one of the most persuasive phrases in higher education. It signals care, equity and institutional responsibility. Increasingly, it also signals dashboards, including AI-enabled systems that aggregate student data, infer risk and surface recommendations to advisors, faculty and administrators.
These systems are often framed as supportive tools or internal analytics. In practice, many function as record-producing infrastructure that reshapes how students are evaluated, supported and sometimes disciplined. When that happens, FERPA is not peripheral to student-success initiatives. It is central to them.
The risk is not that universities want to help students. The risk is that AI-driven dashboards create and redistribute education records without the governance structures that FERPA requires.
Table of Contents
- Student-Success Dashboards Are Not Neutral
- Inference Creates Records — Even When Data Already Exists
- Redisclosure Happens Inside the Institution
- 'De-Identified' Is Often Re-Identifiable
- The Vendor Control Problem
- Leadership Accountability Is Increasing
- Governing Student Success Dashboards Responsibly
- The Line Institutions Must Hold
Student-Success Dashboards Are Not Neutral
Traditional advising relied on discrete records: grades, credits earned, attendance indicators. Student-success dashboards synthesize far more. They ingest LMS activity, assignment metadata, interaction patterns, assessment results and, in some cases, behavioral or engagement signals. Machine learning models then infer likelihoods such as risk of withdrawal, probability of failure or need for intervention, and present these in visual formats designed to prompt action.
These outputs may feel like insights. Legally and institutionally, they are records.
FERPA defines an education record as any record directly related to a student and maintained by an educational agency or a party acting on its behalf. AI-generated risk scores, alerts and recommendations tied to identifiable students meet that definition once they are retained or retrievable. Calling them “analytics” does not change their status.
Empirical research confirms that these inferred records often become more influential than the underlying data, shaping advising and progression decisions across the institution.
Inference Creates Records — Even When Data Already Exists
A common institutional assumption is that dashboards merely analyze existing records and therefore do not create new obligations. That assumption misunderstands how inference works.
FERPA does not distinguish between raw and derived records. When a system uses existing education records to generate a new conclusion, such as a risk score or intervention recommendation. That conclusion is itself directly related to the student and maintained for institutional use. It becomes an education record.
Contemporary learning-analytics research (here and here) shows that inference collapses data into judgment, often without transparency or opportunities for contestation. AI accelerates this transformation and scales it across entire student populations.
The legal implication is straightforward: derived judgments are not exempt from governance simply because they are predictive.
Related Article: The Agentic AI Trap — and the Compliance Line Universities Keep Crossing
Redisclosure Happens Inside the Institution
Student success dashboards are designed to broaden access. A risk flag visible only to an instructor is limited in scope; a dashboard accessible to advising teams, student-support units, administrators and sometimes external partners, is not.
FERPA permits disclosure of education records without consent to school officials with legitimate educational interests, but it also strictly limits redisclosure. As access expands, institutions must be able to demonstrate that each role viewing the data has a documented, policy-defined educational interest. Broad appeals to “student success” are insufficient.
Research (here and here) on learning-analytics governance consistently shows that dashboards stress redisclosure boundaries by design, expanding visibility faster than institutions formalize purpose limitation or access controls.
When AI outputs are widely visible, redisclosure analysis is unavoidable.
'De-Identified' Is Often Re-Identifiable
Vendors and institutions frequently claim FERPA safety through de-identification. In practice, that claim rarely holds in institutional contexts.
Student success dashboards often allow filtering by program, course, cohort or time window — features that make re-identification trivial within a university. Advisors and administrators do not need names to know which student a dashboard is describing.
Research on re-identification demonstrates that contextual data points are often sufficient to re-link records to individuals, even when direct identifiers are removed. As a result, de-identification becomes a technical aspiration rather than a legal safeguard.
The Vendor Control Problem
FERPA allows vendors to access education records only under narrow conditions, most commonly the school-official exception. That exception requires direct institutional control, a legitimate educational interest and enforceable limits on redisclosure.
Student success dashboards complicate this arrangement. When vendors retain analytics, tune models, benchmark across clients or provide “continuous improvement,” institutions often lose the ability to enforce FERPA’s limits. Responsibility does not transfer with the contract. It remains with the university.
Governance research shows that institutions routinely underestimate this shift, accepting sophisticated analytics without ensuring they can meet registrar-level obligations for records they did not previously acknowledge.
Leadership Accountability Is Increasing
FERPA compliance has traditionally been delegated to registrars or compliance offices. AI-driven dashboards change that calculus. When inferred records influence advising, progression or intervention decisions, FERPA becomes inseparable from academic governance and institutional strategy.
Boards and presidents are increasingly accountable for how technology reshapes student outcomes and student rights. When AI-generated records surface during appeals, grievances, accreditation reviews or litigation, leaders must explain not only the decision, but the system that produced it.
AI exposes FERPA risk not because it is harmful, but because it is efficient.
Webinar
Mar
25
Content Leaders Collective: Navigating Content Decisions at Scale
Discover how content leaders are modernizing content operations, avoiding costly missteps and preparing for scale and AI.
Register
Webinar
On demand
Content Strategy Leaders Live: Scaling for Speed, Complexity and AI in High Tech
A candid roundtable on how high-tech leaders are rethinking content at scale.
Watch Now
Webinar
On demand
Do More with Less: Modernizing the Cloud Contact Center for 2026
Learn how to leverage cloud platforms without adding a single hire to personalize every customer interaction.
Watch Now
Webinar
On demand
Cut the Noise: Deploying AI That Actually Moves the Needle
Learn how to turn AI experimentation into concrete revenue operations.
Watch Now
Webinar
On demand
Ditch the Desk Phones: How Modern Teams Drive AI-First Communications
Find out how one team finally pulled the plug on a legacy phone system. And built something smarter.
Watch Now
Webinar
On demand
Rebrand. Migrate. Optimize. How to Do It All (Without Slowing Down)
Cresta leveled up site speed, design flexibility and marketer sanity (in record time). Find out how.
Watch Now
Governing Student Success Dashboards Responsibly
Universities can deploy AI-enabled dashboards responsibly within FERPA’s framework, but only if they impose discipline:
- Define which AI outputs constitute education records
- Limit access by role with documented legitimate educational interests
- Enforce least-privilege access and audit usage
- Prohibit secondary vendor use and set short retention periods
- Treat AI outputs as records subject to inspection and amendment
These are governance choices, not technical limitations.
Related Article: AI in the Wild: Executive Orders Don’t Rewrite FERPA
The Line Institutions Must Hold
Student success dashboards promise clarity through analytics. FERPA demands accountability through control.
When AI systems infer, retain and redistribute student-level judgments, they create education records whether institutions acknowledge them or not. Universities that govern these systems explicitly can pursue student success without eroding student rights. Those that do not will encounter FERPA not during planning, but during conflict.
Student success does not excuse record-keeping. It intensifies it.
Learn how you can join our contributor community.