Responsible AI
AI supports instruction; saved evidence stays authoritative.
IntiQuest uses bounded AI workflows for explanations, tutoring, diagnostics, coaching, and learning reflections.
What AI may do
AI can explain a submitted question, ask a guided follow-up, summarize completed learning activity, and recommend a next focus using supplied evidence.
What AI may not do
AI does not change the stored correct answer, silently edit mastery history, guarantee an SAT score, make admissions decisions, or request sensitive personal information. Failed AI work must not prevent continued practice.
Safety controls
Requests run server-side with structured schemas, timeouts, bounded retries, deterministic fallbacks, prompt versions, pseudonymous identifiers, token telemetry, and provider storage disabled.