Plusieurs incidents présumés de tricherie assistée par IA signalés dans les universités SKY de Corée du Sud pendant les examens de mi-semestre d’octobre 2025

Impact commercial & risques PME

Massive cheating scandals at elite South Korean 'SKY' universities—where students systematically bypassed remote exam restrictions using ChatGPT and private messaging groups—devalued academic credentials, triggered widespread media outrage, and caused a severe loss of employer trust. The universities faced administrative chaos, mass student disciplinary hearings, and extensive expenditures on proctoring upgrades and legal fees to defend grading actions. Regulatory Impact Alignment: AI grading algorithms, automated plagiarism detectors, and remote exam proctoring systems operate in high-risk sectors under EU AI Act Annex III. Educational institutions must provide a formal appeal path, safeguard student FERPA privacy rights, and ensure transparent algorithmic auditing.

Leçon de conformité clé

Static online examinations and standard lock-down browsers are highly vulnerable to generative AI exploitation. Academic institutions must recognize that traditional testing models are obsolete in the era of pervasive LLMs. Evaluation frameworks must adapt to measure critical reasoning rather than rote recall. Compliance Audit Standards: For detailed verification audits, this case maps directly under FERPA Student Privacy Standards & EU AI Act Annex III (High-Risk Classification). Systems deploying similar AI features must maintain dynamic security logs and hold systematic compliance records.

Plan d'action étape par étape

• 1Integrate Advanced Proctoring: Deploy advanced, multi-factor browser lockdown tools integrated with automated, hybrid proctoring audits.
• 2AI-Resistant Exam Redesign: Redesign examination structures to focus on open-book, critical analysis, and verbal defense evaluations that cannot be trivially answered by LLMs.
• 3Automated Exam Log Audits: Conduct regular audits of online exam logs to detect highly anomalous submission patterns or synchronized IP address activities.
• 4Academic Integrity Upgrades: Update academic integrity contracts to explicitly define unauthorized generative AI inputs and enforce strict disciplinary penalties.
• 5Automated Appeal Channel: Establish an active, human-moderated student appeal queue to override false-positive plagiarism or cheating accusations.
• 6Audit Trail Logging: Generate cryptographically signed, tamper-proof logs of all automated grading decisions for FERPA transparency.
• 7Bias Evaluation Audits: Conduct quarterly demographic evaluations to ensure grading software does not discriminate against ESL students.