Research where mistakes are costly
Afritech Labs studies how AI behaves in health, education, security and child safety, and turns what we learn into products and engineering practice.
Research areas
Four areas, one question: can the people affected by this system trust it?
Cybersecurity AI
AI that defends systems, and defences against AI used to attack them.
- How do detection models stay reliable as attackers adapt?
- How do social engineering attacks exploit AI assistants?
Educational AI safety
AI tutors that help students learn without teaching them the wrong things.
- What makes an AI tutor pedagogically safe?
- How should tutoring systems be tested with real students?
Medical AI
Diagnostic and clinical AI that clinicians can check and trust.
- Does a model perform equally well across the populations it serves?
- What evidence should a clinician see before relying on it?
Child safety AI
Protecting children in digital spaces without eroding their privacy.
- How can AI detect risk to children while keeping false alarms low?
- Who should review what the system flags?
Publications
Papers and preprints by our team, kept up to date from Google Scholar.
The Towers Were Standing: A Cause Decomposition of Cellular Outages During Hurricane Helene
DANTE.v2: A Scenario-Specific Breakdown of an LSTM-Based Approach to Predict Insider Threat on Daily System Logs
MC-CPO: Mastery-Conditioned Constrained Policy Optimization for Pedagogically Safe Intelligent Tutoring Systems
Pedagogical Safety in Educational Reinforcement Learning: Formalizing and Detecting Reward Hacking in AI Tutoring Systems
AMDS: Attack-Aware Multi-Stage Defense System for Network Intrusion Detection with Two-Stage Adaptive Weight Learning
Adversarial Machine Learning for Robust Password Strength Estimation
Automatic Detection of Lumbar Spine Disc Herniation: Using Computer Vision and Artificial Intelligence
AI-Powered Instant Textual Feedback on Physiotherapist Student Practical Performance
A keyword-based tool that gives physiotherapy students instant written feedback on their practical examinations.