Securing critical infrastructures is vital to mitigate risks in the rapidly advancing digital era. Manual threat detection during cyberattacks proves challenging due to human effort and time constraints. Integrating AI-based threat detection offers a robust solution, leveraging AI to recognize, categorize, and mitigate cyberattack repercussions. Over the past five years, identifying suitable machine learning and AI algorithms for scrutinizing threats in critical infrastructure protection has become a central concern. Furthermore, how can threat detection seamlessly integrate into essential infrastructure cybersecurity, emphasizing AI? This research proposes a Supervised Learning model known as a Decision tree for threat detection. The method involves accumulating data from the NSL-KDD vulnerability database and meticulous preprocessing. Refined data undergoes analysis using a Decision Tree model, adept at recognizing existing vulnerabilities and predicting new ones. The research showcases less than 10% false positive rate for each attack category and an impressive accuracy score of 99.76%. These outcomes validate the efficacy of this research in achieving a robust and accurate cybersecurity model. Effective threat detection is pivotal for an in-depth cybersecurity strategy, enabling proactive identification and mitigation and minimizing potential cyberattack impact.