Prototype. Full-site concept for civitaslearning.com. URL structure matches the live site. Figures marked "example" are illustrative.v0.3 · Astro · Sept 2026
Home / Company / Responsible AI
Responsible AI

AI with student data has to earn trust. Here is how ours does.

Fifteen years of building AI for higher education taught us what responsible looks like in practice: your students’ data, a person in the decision, and a record of everything the AI did.

Our principles

Six commitments for AI that touches student records.

Each one is already in how the platform works today, not a policy to come.

01

Trained only on your students

Every model is built on your institution’s own history. Your data is never pooled into a benchmark sold to another school, and it is never used to train models for another institution.

02

A person keeps the decision

AI recommends and drafts; staff set the audience, the timing, and the message, and approve the action. In the Student Success Impact Report, institutions that deployed AI without human involvement saw persistence fall 6.8 percentage points (observational analysis in the Student Success Impact Report; sample size not published).

03

Every action is on the record

Every AI action is logged, attributed, and reviewable. Your staff set the guardrails, and the audit trail belongs to the institution.

04

Predictions you can explain

Each prediction shows the factors that drove it, so an advisor sees the reason, not just a score.

05

Models you can audit

Models are documented and retrained every term. Institutional research can examine model behavior by student population, instead of trusting a single accuracy figure.

06

Measured, not assumed

Every initiative is measured against a matched comparison group, with the population size and the confidence interval — including the ones that produced nothing.

For students

The same rules on the student’s phone.

AI that talks to students carries more responsibility, not less.

Only what a student may seeIn Inspire, the student app, the assistant reads only records the student can already open, and a student sees only the notes the support team chose to share.
Advisors stay in the loopThe assistant tells students to confirm important academic decisions with an advisor.
The foundation

Your data stays yours.

The documents behind these commitments are published. Security documentation is available through the Trust Center.

FERPA school official

Civitas works as a school official your institution designates, under a published data sharing agreement.

Data sharing agreement →

Access follows the role

Single sign-on and role-based access: what a person can see follows the role the institution assigns.

Platform specifications →

Not for sale

We do not sell data to third parties, and we protect it with industry best practices, including encryption and de-identification.

Privacy policy →
Security documentation

Review our security posture and request documentation in the Civitas Learning Trust Center →

Frequently asked questions

Questions institutions ask about responsible AI

Short answers, in the words a cabinet uses. For anything not here, ask us directly.

Is our student data used to train AI for other institutions?

No. Every model is trained on your institution’s own history, and your data is never pooled into a benchmark sold to another school.

Does the AI make decisions about students on its own?

No. AI recommends and drafts; staff set the audience, timing, and message and approve the action. In the Student Success Impact Report, institutions that deployed AI without human involvement saw persistence fall 6.8 percentage points (observational analysis in the Student Success Impact Report; sample size not published).

Can we see what the AI did on our behalf?

Yes. Every AI action is logged, attributed, and reviewable, and the audit trail belongs to the institution.

Can our institutional research office audit the models?

Yes. Models are documented and retrained every term, each prediction shows the factors behind it, and IR can examine model behavior by student population.

Where do we get security documentation?

In the Civitas Learning Trust Center, where you can review our security posture and request documentation.

Next step

Bring your questions. Leave with an estimate you can defend.

A 45-minute working session: we start from your public IPEDS figures, show what the platform found at institutions like yours, and size the outcome — before anyone touches your data.