Cyber threats are constantly evolving due to the increased reliance enterprises have on systems and networks to conduct their operations. Malicious actors, such as Advanced Persistent Threat (APT) groups, are among the most significant dangers due to their sophistication and relentless exploitation of vulnerabilities. To further complicate this issue, most often, APT attacks are complex and challenging to discover. This results in enterprises struggling to identify the actual perpetrator behind an attack as well as the appearance of their tactics. This research not only emphasizes the importance of up-to-date cyber threat intelligence but provides insight into methodologies on how these attackers’ true identities can be exposed. This offers leverage to enterprises to uncover who is attacking them and what their attack patterns are, thus providing better mitigation against further damage and exposure. This dissertation explores seven different machine learning algorithms to detect unknown APT groups from within a dataset which contributes significantly to cyber threat intelligence and bolsters enterprise security. Furthermore, it explores successive dataset versions to provide evidence of performance improvements in the selected algorithm, underlining the necessity of the latest threat intelligence on APTs to improve defensive postures. Finally, to give a more comprehensive overview of how the threat landscape is shifting and evolving, a comparative analysis of APT tactics, techniques, and procedures (TTPs) is explored so that enterprises can stay ahead of the evolution of strategies employed by these groups. This is conducted using successive and current dataset versions to examine repeated patterns APTs use and what emerging threats enterprises should be most concerned about to build more robust defense strategies. As the world of cyber threats evolves, adaptive threat intelligence datasets and methodologies concerning APT groups greatly benefit enterprises as these attacks will remain a persistent threat.