Fred: A Campus Assistant That Won’t Promise Jobs or Reveal Records

A prospective student asks your institution’s website how much aid they will get and whether graduates find jobs. A general-purpose chatbot estimates an aid package and assures them graduates almost always land work. Neither number came from your financial aid office or your outcomes data. The institution wanted a tool to help applicants. It got one making an individualized aid determination and an employment promise that federal rules treat seriously.

The threat and standard articles in this series walk through how those claims become misrepresentations. This one is the answer: what a compliant AI assistant for higher education looks like, and why the design, not a disclaimer, is what keeps a website from promising outcomes, quoting aid, or disclosing records.

The Real Choice Is Governed or Ungoverned

Institutions ask whether to put AI on the site at all. Students already decided. They research at all hours and want fast answers. The real question is whether the assistant is governed.

An ungoverned chatbot promises jobs, estimates aid, and reads back records to anyone. A governed one explains programs and how aid generally works, captures inquiries, and routes outcomes, aid figures, and records to the right office. Same convenience, completely different exposure when a misrepresentation or a FERPA release surfaces.

How Fred Holds the Line

Fred starts from the opposite default of a general chatbot. A generic bot answers anything unless told not to. Fred answers only what it is cleared to, and the boundary lives in the system rather than a prompt that drifts when a student pushes.

In practice, Fred works from your own content. It explains programs, deadlines, how to apply, and how aid generally works. It does not guarantee jobs or outcomes or quote an individualized aid figure, because the Department of Education’s rules on substantial misrepresentation reach exactly those claims, alongside the FTC’s deceptive-practices authority. It does not disclose student records to an unverified person, because the Family Educational Rights and Privacy Act protects them, and it answers accessibility questions accurately under Section 504. Fred runs more than 50 industry guardrail packs, and the higher-education pack is built around the misrepresentation rules, FERPA, and Section 504.

The design proves itself under pressure. A student will ask "will this actually get me hired?" several ways, and a prompt-instructed bot eventually reassures them. Fred does not rely on catching the phrasing. It decides what may be said before the reply forms, so an outcome promise or an aid figure never reaches a student. "Will not" is a suggestion. "Cannot" is an architecture.

Fred vs. a Generic AI Chatbot

That single screen is the argument. A general chatbot is encouraging until encouraging becomes a misrepresentation or a record disclosure. Fred is helpful across everything that moves a student toward applying and structurally incapable of promising outcomes or revealing records.

What Your Institution Actually Gets

Set the rules aside and look at admissions. Fred handles the after-hours questions that usually wait, what programs do you offer, when is the deadline, how do I apply, how does aid generally work, and answers them instantly. It captures inquiries with the details your team needs, and routes outcomes, aid figures, records, and accommodations to the proper office.

The captured-applicant math is the real win. The student who gets a fast, accurate answer at 10 p.m. is the one who applies, instead of the one who drifted to another school. Fred turns the after-hours window into applications without ever putting an outcome promise or a record disclosure on the record.

So the question is not whether institutions will use AI. They will. It is whether yours routes outcomes, aid, and records to the right office, every time, no matter how the question is asked.

Frequently asked questions