Academic integrity and appeals

Ogov checks submitted work for source integrity. A similarity score measures matching text, not misconduct: quotations, legislation, references and common policy language all produce matches. Ogov therefore never publishes a similarity or AI-detection percentage, and never states one about a named person.

If a concern is raised

• An assessor examines the actual matches, not the number
• The learner is shown the concern and may explain or correct
• A separate academic-integrity panel decides, and the assessor who raised it does not sit on that decision
• The learner may appeal to a differently composed panel
• Only confirmed serious misconduct affects a credential

A confirmed revocation is public on the credential's verification page, because anyone relying on that credential deserves to know. The evidence and reasoning stay private to the parties.

Authorship is established by evidence rather than software: version history, a source and evidence log, contributor and AI-use declarations, and a live defence in which the author answers unseen questions about assumptions, trade-offs and rejected alternatives.