By Rita Okoye

Cloud computing has become an essential part of modern digital infrastructure, powering everything from financial transactions to healthcare systems and global communication networks. Its ability to provide on-demand access to data, scalable resources, and cost-effective solutions has made it the backbone of industries worldwide.

However, this digital revolution has also given rise to sophisticated cyber threats, particularly Distributed Denial-of-Service (DDoS) attacks. These attacks target cloud-based services by overwhelming them with massive amounts of malicious traffic, disrupting normal operations and causing millions in financial losses.

For Nigerian cybersecurity expert Chisom Elizabeth Alozie, tackling this challenge became the focus of her MSc research at Robert Gordon University (RGU) in Scotland.

 Her groundbreaking study, “Detection of Cloud DDoS Attacks Using Supervised Machine Learning,” explores how artificial intelligence can revolutionize cloud security by identifying and neutralizing cyber threats before they escalate.

Alozie’s research is a timely contribution to the ongoing global effort to fortify cloud infrastructure against cybercriminals who constantly evolve their attack methods.

Speaking on the urgency of her study, Alozie states, “Cloud computing is the future, but without strong security measures, it remains highly vulnerable. The ability of cybercriminals to disrupt entire networks through DDoS attacks is a major concern for businesses, governments, and individuals. My research aims to provide a proactive, AI-driven solution to this growing problem.”

At RGU Scotland, Alozie conducted an extensive study into how supervised machine learning can be leveraged to detect and mitigate DDoS attacks in real-time cloud environments. Her research methodology involved creating a controlled cloud-based system to simulate both benign and malicious traffic, collecting data from these interactions, and applying various machine learning algorithms to analyze attack patterns. By doing so, she was able to determine which models were most effective at distinguishing between normal network activity and DDoS attack traffic.

“Traditional security measures are no longer sufficient,” Alozie explains. “Firewalls and signature-based detection systems struggle to keep up with rapidly evolving cyber threats. What we need is a more intelligent approach—one that can learn from past attacks and anticipate future ones before they cause damage.”

Her study specifically tested four machine learning algorithms—Random Forest, Support Vector Machine (SVM), Naïve Bayes, Decision Tree, and K-Nearest Neighbors (KNN)—to evaluate their effectiveness in detecting DDoS attacks. By applying these models to both a newly generated dataset and an open-source dataset (CSE-CIC-IDS2018), Alozie was able to compare performance metrics such as accuracy, precision, recall, and F1-score. The results were striking: Random Forest, Decision Tree, and KNN emerged as the top-performing models, achieving over 99% accuracy in detecting malicious traffic.

Explaining the significance of these findings, Alozie says, “What this research shows is that AI can be a game-changer in cybersecurity. With the right machine learning model, we can drastically reduce false positives and enhance threat detection capabilities, making cloud environments much safer for users.”

However, deploying machine learning for real-world cloud security solutions is not without challenges. One major concern highlighted in her research is scalability—the ability of machine learning models to handle increasing amounts of network traffic as cloud systems expand. Alozie emphasizes the importance of continuous optimization and retraining of AI models to ensure they remain effective under different network conditions.

“The biggest challenge with AI-driven security is ensuring that models can scale,” Alozie explains. “A detection system that works well in a lab environment might struggle when deployed across a large-scale cloud network. That’s why ongoing research and optimization are crucial.”

Another key challenge is adversarial attacks, where cybercriminals manipulate network traffic to bypass detection systems. Alozie’s research underscores the need for resilient AI models that can adapt to evolving attack strategies. Her findings recommend that cloud security teams regularly update their machine learning models to counteract new evasion techniques employed by cybercriminals.

Despite these challenges, Alozie remains optimistic about the future of AI in cybersecurity. She advocates for a hybrid approach that combines machine learning with traditional security measures, ensuring a multi-layered defense system capable of detecting and mitigating threats at various levels.

“No single security solution is foolproof,” Alozie acknowledges. “AI-driven detection should complement existing security frameworks, including firewalls, encryption, and behavior-based anomaly detection. The goal is to build a comprehensive defense strategy that leaves no room for attackers to exploit vulnerabilities.”

Beyond the technical implications, her research also explores the legal, ethical, and corporate responsibility aspects of AI-powered security. In an era where data privacy regulations such as the GDPR (General Data Protection Regulation) are becoming stricter, cloud security solutions must balance intrusion detection with user privacy. Alozie stresses the importance of transparent AI development, ethical data handling, and compliance with international cybersecurity laws.

“AI in cybersecurity must be implemented responsibly,” she asserts. “While it is essential for preventing attacks, we must ensure that user privacy is not compromised in the process. Striking the right balance between security and ethical AI usage is critical for long-term success.”

Her study also highlights the role of corporate social responsibility (CSR) in cybersecurity, emphasizing that businesses must invest in proactive security measures not only to protect themselves but also their customers, employees, and stakeholders. She argues that cloud service providers must actively engage in security research, collaborate with cybersecurity experts, and share intelligence on emerging threats.

“Cloud security is a shared responsibility,” Alozie explains. “Service providers, enterprises, regulators, and security researchers must work together to create a safer digital ecosystem. The days of reactive security are over—we need to be proactive, predictive, and preventative.”

As she concludes her research at RGU Scotland, Alozie remains committed to pioneering new AI-driven solutions for cloud security. She envisions a future where cloud-based AI systems can autonomously detect, analyze, and neutralize cyber threats in real time, providing businesses and individuals with a more secure digital environment.

“Cyber threats are evolving, but so is technology,” she concludes. “By integrating AI, machine learning, and real-time data analytics, we can develop security solutions that are smarter, faster, and more adaptive than ever before. The future of cybersecurity lies in AI-driven intelligence, and I am proud to contribute to this evolving field.”

Her research stands as a testament to the potential of AI in defending cloud environments against cyber threats. As cloud adoption continues to grow, Alozie’s work provides a roadmap for how organizations can implement cutting-edge machine learning techniques to protect their digital assets from malicious attacks. Her contributions to the cybersecurity field at RGU Scotland not only enhance the global discourse on AI in security but also highlight the role of Nigerian experts in shaping the future of cloud defense strategies.