JAMB UTME 2025 — AI-Enabled Exam Fraud and the Limits of Biometric Integrity Systems

The Joint Admissions and Matriculation Board (JAMB) administers Nigeria's Unified Tertiary Matriculation Examination (UTME), a computer-based test taken by well over a million candidates a year, using fingerprint and facial-recognition verification at registration and test centres since around 2015. In 2025, following a major cheating and technical-glitch scandal, JAMB convened a Special Committee on Examination Infractions, chaired by Prof./Barr. Jake Epelle, to investigate systemic integrity failures in that year's UTME cycle.

The committee's findings, reported by multiple independent Nigerian news outlets, documented over 6,000 technology-enabled malpractice cases, including 4,251 cases of "finger-blending" (deliberate fingerprint manipulation to defeat biometric identity checks), 190 cases of AI-generated image morphing used to impersonate candidates during facial verification, and 1,878 false disability declarations used to obtain exam accommodations. The panel recommended cancelling the results of 6,319 candidates and described the fraud as "highly organised, technology-driven, and culturally normalised."

Rather than showcasing a working AI solution, this case documents AI being weaponised against an existing integrity system, and a regulator's response: the committee explicitly recommended that JAMB deploy AI-powered biometric anomaly detection, real-time monitoring, and a national Examination Security Operations Centre going forward, implying that robust AI-based countermeasures were not yet in place. It is included here as an honest, well-documented cautionary case about the two-sided nature of AI in high-stakes assessment integrity, rather than a success story.

Read the full analysis: https://punchng.com/jamb-panel-uncovers-over-6000-ai-biometric-malpractices/