How Divine Iloh is advancing applied AI across security, privacy, and resilient infrastructure

By Tosin Clegg

As artificial intelligence systems increasingly power everything from educational platforms to cybersecurity defenses, the question of trust has become paramount. For Divine Iloh, a data leader and artificial intelligence researcher based in the United States, this challenge represents the defining frontier of modern AI development. The stakes are high: AI deployed in critical domains like education, healthcare, financial services, and cybersecurity cannot afford to fail, compromise privacy, or collapse under real-world constraints. Yet building systems that are both powerful and trustworthy requires navigating complex tradeoffs between performance, transparency, and resilience.

Speaking Iloh explained: “AI is moving into domains where the cost of failure is high.” “The work now is to build models and products that perform well, while also meeting requirements like reproducibility, transparency, and resilience.”
This philosophy has guided Iloh’s trajectory from Senior Analyst to Data Analytics Manager at Walmart Inc., where he now spearheads enterprise analytics, AI integration, and automation initiatives, including forecasting and standardizing AI-driven systems that drive critical decision-making. His earlier role involved building scalable data pipelines and machine learning models that delivered operational and risk insights across the organisation.

But Iloh’s impact extends far beyond corporate halls. In 2025 alone, he has made significant contributions to academic research that could reshape how institutions handle sensitive data and cybersecurity threats.

His publication “Generative Private Synthetic Student Data for Learning Analytics: An Empirical Study” in IEEE Access examines how deep generative models can create synthetic educational datasets while minimizing privacy risks. The study evaluates both statistical and deep generative methods, demonstrating how synthetic data can advance learning analytics research without compromising sensitive student records.

He also co-authored “An Optimized Deep Learning Framework for Malware Classification Using Integrated LSTM and CNN Approaches,” published in the International Journal of Advances in Engineering and Management. The research presents an optimized malware-classification framework centered on an integrated CNN-LSTM architecture, with detailed model design and evaluation using standard classification metrics.
Iloh’s innovative spirit has also manifested in intellectual property that addresses real-world challenges. His granted patent, “Bandwidth Aware, Curriculum Aligned Multi-Tenant System for Accredited Remote Education with Offline Assessment Integrity,” describes an AI-driven, connectivity-aware system designed to enable remote learning in low-bandwidth, intermittent connectivity environments, a critical need in underserved communities globally.
As co-inventor on “AI-Driven Cybersecurity & Anomaly Detection Framework for IoT Supply Chain,” he has developed an AI-based framework capable of detecting anomalies in near real time and triggering defensive responses across multiple sectors.

“The technical challenge is not only learning from data,” Iloh noted. “It’s designing systems that remain dependable when the data, the environment, and the threat model change.”

This commitment to building trustworthy AI in production environments, whether through privacy-preserving data methods, robust detection for adversarial environments, or architectures that operate reliably despite constraints, defines Iloh’s body of work.
In October, he took his message to a broader audience, delivering a keynote speech at the 2025 Nigeria Innovation Summit. His address focused on approaches to building AI-ready institutions and practical pathways for adopting AI responsibly in emerging markets, emphasizing implementation, governance, and measurable outcomes.

As he looks to the future, Iloh plans to expand his research on privacy-preserving learning and AI-assisted monitoring systems that can scale across industries while maintaining transparency and auditability, continuing his mission to make AI both powerful and trustworthy.