*Team’s breakthrough work on securing Internet of Things networks earns top honours at ACSCON-2026 in Bengaluru A team of six Nigerian researchers has been honoured with the Best Paper Award at the International Conference on Advances in Computing Systems, Communication and Networking (ACSCON-2026), held on 29th and 30th May 2026 at Dr. ACS College of Engineering, Bengaluru, India. The award-winning paper, titled “Federated Learning with Differential Privacy for Adversarial Robustness in IoT Cybersecurity: Defending Against Data Poisoning Attacks,” was authored by Babatunde Alexander Abiola, John Olalere Ogunlola, Opeyemi Victor Omolade, Blessing Alice Alao-Olatunji, Adedeji Edward Adesola, and Simon Olufikayo Awodele. Five of the authors; Abiola, Ogunlola, Omolade, Alao-Olatunji and Adesola are researchers at the Research and Doctoral College of the University of Greater Manchester, Bolton, United Kingdom, while Awodele is of the Department of Computer Science, Babcock University, Ilishan-Remo, Ogun State, Nigeria. The collaboration underscores the growing contribution of Nigerian scholars, both at home and in the diaspora, to cutting-edge research in artificial intelligence and cybersecurity. Tackling a Global Security Challenge The research addresses one of the most pressing challenges in modern cybersecurity: how to protect the billions of Internet of Things (IoT) devices from smart home appliances to hospital sensors and industrial equipment from cyberattacks, without compromising the privacy of the data those devices generate. The team’s solution combines two powerful techniques. Federated learning allows many devices to jointly train an artificial intelligence security model without ever sharing their raw data, while differential privacy adds mathematical guarantees that no individual’s information can be extracted from the shared model. The researchers then went a step further, stress-testing their system against sophisticated “data poisoning” attacks a technique in which hackers secretly corrupt the training process to sabotage the security model from within. Their results were striking. Using real-world IoT data, the