Phishing remains a pervasive threat in the online landscape, challenging internet users' security. This study conducted a Systematic Literature Review (SLR), analyzing 80 articles to understand anti-phishing strategies. Spear phishing, email spoofing, email manipulation, and voice phishing were identified as common techniques. Machine learning emerged as a promising solution for combating phishing attacks.The research focuses on using Logistic Regression and Multinomial Naïve Bayes algorithms for phishing detection. Logistic Regression demonstrated superior accuracy at 96.3%, surpassing Multinomial Naïve Bayes by 1%. Ensemble learning, weighted soft voting, and Kappa statistics were employed to enhance classification, achieving 96.3% accuracy. This work underscores the effectiveness of ensemble learning in phishing detection, suggesting its potential to bolster internet security. Future research directions include exploring various phishing methods and anti-phishing measures for enhanced protection.