TAISE case studies
TAISE · Put your knowledge to work in context.
This independent study page provides practice for TAISE. Its editorial learning areas include AI foundations, data and deployed-system lifecycle; AI threat boundaries, agents and technical safeguards; AI risk, accountability and responsible use; Identity, access and cryptographic protection; Security operations, monitoring and response; Cyber-risk identification and analysis; Continuity, backup and technical recovery. The TAISE case studies are 5 cases and about 1,052 words.
1Retrieved text asks an assistant to exceed its authority
A fictional AI application retrieves a document to summarize it for a user. The document contains text instructing the assistant to ignore its task and send private records to a new external address.
183 words, under 1 min2SaaS access after a role change
A fictional organization uses a managed collaboration service for internal planning documents. The provider maintains the application and underlying servers, while the organization configures users, groups and sharing settings.
221 words, about 1 min3An incident review with an action that can be checked
A fictional delivery team restores service after a release encounters an undocumented database dependency. The initial meeting ends with the statement that the engineer should be more careful, but no owner, test or review date is recorded.
219 words, about 1 min4A control comparison with explicit effectiveness and uncertainty
A fictional organization estimates the value exposed by one defined incident scenario at five hundred thousand currency units and the fraction lost in one successful incident at forty percent. Its classroom model estimates an annualized incident rate of one half, and the exercise asks for expected annual loss rather than a prediction that exactly half an incident occurs each year.
206 words, about 1 min5A recovery target does not prescribe one universal architecture
A fictional order service has a recovery-point objective of fifteen minutes and a recovery-time objective of four hours. A tested option creates consistent recovery points every ten minutes and restores the service in three hours after the declared recovery procedure begins.
223 words, about 1 min