Academic Integrity and AI Policy

ACADEMIC INTEGRITY AND AI POLICY

Governance of similarity, AI indicators, AI tutoring and academic decision support

Academic Quality. Verified.

1. Core principle

2. Covered functions

• Similarity/originality and cross-student comparison where lawfully configured.
• AI-content indicators.
• Grammar, writing-quality, citation/reference and source validation.
• Approximate marks/readiness indicators.
• AI Tutor and learning diagnostics.
• Handwriting/verification features where separately enabled.

3. Human review and challenge

Where a result may materially affect a person, the affected user should have a visible route to provide context, challenge the result and obtain qualified human review. Institutions remain responsible for disciplinary procedure and final decisions.

4. Similarity and repositories

Veri-AQ may compare work against authorised public sources, approved scholarly sources, the user’s own prior work and institution-specific repositories where lawfully enabled. Veri-AQ does not currently operate a global cross-customer repository of all user submissions or derived fingerprints. Institution repositories require documented purpose, authority, notice, access control, retention and exit/deletion rules.

5. Model training

Veri-AQ does not currently use user assignments, prompts, AI Tutor conversations, reports, learner content or user histories to train general-purpose or Veri-AQ AI models. If model training using personal information or user content is proposed in future, it is a new material purpose requiring legal/privacy assessment, retention rules, transparency and any required separate consent before activation.

6. AI Tutor integrity

• Provide explanation, scaffolding, practice and feedback rather than facilitating academic fraud.
• Use academic-integrity gates to block substantive completion of graded work where prohibited.
• Minimise user content before any approved external AI request.
• Do not send unrelated billing, family, other learner or account-secret information.

7. Children

Child-facing tools require age-appropriate explanations, active guardian/competent-person controls where applicable and negative testing to prove that private child content cannot leak to analytics, logs or unapproved providers. U.S. users under 13 require the separately activated COPPA workflow.

8. Validation and model register

• Record model/rules version, purpose, inputs/outputs, hosting/vendor, processing location, retention and known limitations.
• Track material false-positive/false-negative patterns.
• Re-test after material model, provider, rule or hosting changes.

Legal review and mandatory-law notice. This policy framework is designed for Veri-AQ operations and must be reviewed by appropriately qualified counsel before a new entity, jurisdiction, school contract, under-13 U.S. workflow or materially new processing purpose is launched. Nothing in this document excludes rights or duties that applicable law does not permit the parties to waive.

Primary legal references

• COPPA: Children’s Online Privacy Protection Rule, 16 CFR Part 312 — https://www.ftc.gov/legal-library/browse/rules/childrens-online-privacy-protection-rule-coppa
• FERPA: Family Educational Rights and Privacy Act, 34 CFR Part 99 — https://studentprivacy.ed.gov/ferpa
• POPIA: Protection of Personal Information Act 4 of 2013 — https://www.gov.za/documents/protection-personal-information-act

Contact details

Website: www.veri-aq.com | Support: support@veri-aq.com | Privacy / PAIA: privacy@veri-aq.com

Principal address: Hofmeyer Rd, Centurion, South Africa, 0157