Oxonia Policy
AI, governed.
Regulator-ready.
Policy, register, assessment, training and reporting — with one readiness score, and every requirement traced to its source.
Demo University against a DEMO requirement pack. Real packs cite official source documents clause by clause.
Policy builder
A complete policy set, cited clause by clause.
Answer a short institution profile. The Policy Drafter produces eight policies — from the Institutional AI Policy to the AI Governance Committee's terms of reference — and every clause names the requirement it satisfies. Gaps are shown, not hidden.
- Approved through Oxonia Senate — the resolution number is stamped on the policy
- Versions, comparisons, review dates and owners
- Attestation campaigns tracked by faculty
AI-detection scores are low-weight indicators only. They may never on their own be the basis for an academic misconduct case. Process evidence and declarations come first.
AI register
Every AI tool, known and owned.
One register of every AI system your university uses — owner, purpose, data categories, hosting, contract, risk tier, DPIA. Anyone can request a new tool from the AI Hub; the Register Agent researches the vendor's public terms and proposes a risk tier for the committee.
- Shadow-AI discovery from your SSO and LTI app lists — never from anyone's browsing
- Vendor questionnaires with gaps flagged
- Register your own agents and MCP tools with permissions
Processes research-participant audio. Vendor states customer audio is not used for training. Proposed tier: High — DPIA required.
AI in assessment
Evidence, not accusations.
Every assessment declares an AI-use level. Students declare what they used at submission. An optional composition space shows the writing process — not an “AI score”. Detection results, if you license them, are a weak signal that can never open a case alone.
- Assessment Audit agent flags missing or unclear declarations
- Redesign suggestions that hold up at each level
- Fair-process integrity cases with evidence packs and appeals
- Oxonia Mark enforces the level on exam papers
AI may give feedback or edit your own writing. It may not write new content.
Our own wording — every institution can edit it.
Training
Five-minute lessons. On any phone.
The Course Builder turns your approved policies into short lessons, quizzes and scenarios — localised, offline-ready in the AI Hub, or delivered one lesson a day on WhatsApp. Mandatory assignments by role, verifiable certificates, completion by faculty.
- AI literacy for students, staff, researchers and governors
- English, French, Arabic, Portuguese, Hausa, Yoruba, Igbo, Swahili
- QR-verifiable certificates
Regulator report
The report writes itself. Senate signs it off.
The Regulator Report agent drafts your periodic report from live data — policies, minutes, training, register, declarations, incidents — citing the evidence behind every statement. It goes through Senate before it's submitted.
- Readiness score per requirement: met, partial, missing
- Evidence library with expiry reminders
- Time-boxed audit room for external reviewers
- AI incident reporting and triage with DPO escalation
2. Governance. The University constituted an AI Governance Committee (terms of reference approved SEN/2025/019) and approved its Institutional AI Policy SEN/2026/045.
3. Literacy. 694 of 958 staff (72%) have completed AI-literacy training TRN-2026-Q3.
Federation view
For regulators: the whole sector, anonymised.
Federations and regulators see aggregate readiness, policy adoption, training bands and incident categories across member universities — and can publish guidance and run surveys. They never see a university's internal documents unless it chooses to share a report.
- Aggregates only, by design
- Sector guidance pushed to every member
- Surveys with response tracking
How we compare.
Enterprise AI-governance tools govern models. Oxonia Policy governs a university.
| Capability | Oxonia Policy | Credo AI | OneTrust AI Gov. | Turnitin Clarity |
|---|---|---|---|---|
| AI register / inventory | — | |||
| Regulations turned into requirement packs | — | |||
| Evidence recording for audit | — | — | ||
| University policy set drafted by agents | — | — | — | |
| AI-use rules per assessment | — | — | AI assistant per assignment | |
| Student AI-use declarations | — | — | — | |
| Writing-process transparency | — | — | ||
| Staff and student AI training with certificates | — | — | — | |
| Policies approved via Senate minutes | — | — | — | |
| Regulator / federation sector view | — | — | — |
✓ = advertised on the vendor's public materials. — = not advertised (not necessarily absent). Based on public information reviewed in September 2026; tell us if something has changed.
Policy plan
from ₦1.5m / month
$1,200 · £950 — per month, by student band.