Misinformation spreads fast online. I wanted to build a model that could flag suspicious articles automatically — no manual fact-checking required. I tested both Logistic Regression and Naive Bayes.