Educators run more proctored AI-free assessments conditionally on compromised outcomes, and reject AI detectors based on classroom experience of false positives¶

Claim · All claims
Evidence · 1 study (2 entries) · qualitative r2 · q1
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? Participants apply a conditional rule: where a learning outcome is compromised by AI use, that outcome moves to a proctored AI-free assessment, weighted most heavily early in a course. → Muhammad Sajjad Akbar 2026 q1 i? AI detectors were ruled out because participants who had used them reported many false positives, and AI-generated work is often indistinguishable from human work. → Muhammad Sajjad Akbar 2026 (2)

Evidence¶

Muhammad Sajjad Akbar 2026¶

Muhammad Sajjad Akbar, Geoffrey Challen, Fraida Fund, Casey Hopkins, Oscar Karnalim, Kevin Lin, James McGuffee, Shubbhi Taneja, Ranysha Ware. (2026). AI Can Do Your Homework. Now What? Report from an online workshop on computing assessment in the age of generative AI. https://arxiv.org/abs/2609.27842 doi:10.48550/arxiv.2609.27842

q1 · i? · qualitative · r2

Controlled & AI-Free Assessment room, reporting participants' own practice in both small and large classes. The report states no enrollment figures or outcome data were collected, so claims concern what instructors do and believe works.

"The operating rule several described is conditional: where a learning outcome is compromised by AI use, move that outcome to a proctored AI-free assessment."

Muhammad Sajjad Akbar 2026 (2)¶

Muhammad Sajjad Akbar, Geoffrey Challen, Fraida Fund, Casey Hopkins, Oscar Karnalim, Kevin Lin, James McGuffee, Shubbhi Taneja, Ranysha Ware. (2026). AI Can Do Your Homework. Now What? Report from an online workshop on computing assessment in the age of generative AI. https://arxiv.org/abs/2609.27842

q1 · i? · qualitative · r2

Same room, on ensuring a 200-student assignment is AI-free: detectors rejected on experiential grounds; remaining mitigations are dividing verification across tutors, automated marking where possible, and limiting use of expensive methods.

"AI detectors were ruled out on the basis of classroom experience rather than principle: participants who had used them reported many false positives, and AI-generated work is often indistinguishable from human work."

Discussion¶

Related Claims¶

• AI-detection tools produce inequitable outcomes, disadvantaging non-native English speakers through false positives — related
• GPT detectors frequently misclassify non-native English writing as AI-generated, raising fairness concerns for AI policy — related
• In a course that had removed examinations, most student groups designing a fair grading policy reinvented proctored or handwritten assessment
• In AI-assisted software engineering courses, AI-required and programming work carry high median grade weights while proctored assessment is rare — related
• AI text detectors can return high AI-detection scores for human-written academic text, raising fairness concerns about authenticity evaluation — related
• Universities disabling AI detection cite false-positive risk at scale, and detector influence persists in individual grading, journal screening, and institutions without policy revision — related
• A detector can achieve high aggregate accuracy while producing materially different over-detection rates across author groups — related
• In a documented exploratory audit, one human-authored manuscript received materially different classifications from five commercial detectors, spanning 0% human to Human Generated — related
• AI-text detectors misclassify authentic TOEFL essays by non-native English speakers at a mean false-positive rate of 61.3% across seven detectors, far more often than native-speaker essays — related