The 88% Problem: Why Traditional Assessment is Broken in the AI Era
Read more
Post‑submission verification that scales: Students demonstrate understanding of their submitted work through brief recorded reflections.
Universities spend hours conducting in‑person Q&A verification for each suspected AI case, costing time, resources, and money. Instructors suspect AI use but lack efficient ways to quickly verify authentic understanding.
According to the Digital Education Council Global AI Faculty Survey 2025, 54% of faculty believe that current student assessments require significant change. Source (PDF)
Over 50% of students say over‑reliance on AI harms performance and reduces the value they receive. Source
50% of faculty members say assignments should be redesigned to be more AI‑resistant. Source (PDF)
Brief, asynchronous reflections let students evidence their own understanding. No new assignment—just authenticate what was already submitted. Here’s how it fits where you grade:
During grading, instantly flag any submission for understanding checks with one click
Brief, asynchronous authentication of work already submitted — not a new assignment
Transcripts attach directly to original submissions for a complete verification record
Vivas raise scores, fast. In a UC San Diego trial, students who took a 15-minute oral exam with the instructor scored 14% higher on the next written test; motivation also rose, especially for first-generation students.
UC San Diego TodayHEPI/Kortext 2025: usage jumped from 66%→92% in one year; 88% have used GenAI for assessments
HEPI/Kortext Survey (PDF)Study of 666 participants found heavy reliance on assistants is strongly negatively correlated with critical-thinking scores (r ≈ −0.68), mediated by cognitive offloading
MDPI StudyFlag any submission for verification while grading — no pre‑planning required.
Flag submissions for verification during normal grading. Students record a brief reflection; transcripts auto‑attach to the original submission in your LMS. No extra tools, no meetings.
Students explain their reasoning in their own words. You get the authentic voice of the student, not ChatGPT.
VivaEdu solves the immediate AI integrity crisis and it improves speech confidence in students, developing better communicators and increasing future employability.
VivaEdu guides students through brief, contextual reflections on their submitted work to validate understanding.
Flag suspicious submissions during normal grading. No need to pre‑plan or restructure assessments. VivaEdu fits into your academic misconduct procedures, not the other way around.
Canvas • Blackboard • Moodle (LTI 1.3 compliant)
Instructors, educators, and TAs can also sign in to the web app for accessible management and review outside the LMS.
Optional, giving educators full control over additional checkers
Read more
Read more
Read more
Read more
Interested in piloting VivaEdu? Reach out to join our early‑access schools and universities.Reach out to usfor the 2025–2026 academic year. The pilot program is free.
GDPR and DPA‑aligned. LTI 1.3 with scoped tokens.