The expansion of the internet and social media has transformed the way society accesses information, also facilitating the spread of fake news. In the Peruvian context, where linguistic and cultural patterns present unique characteristics, addressing this issue becomes a significant challenge. This work develops a model for fake news detection based on machine learning algorithms, specifically Naive Bayes and Support Vector Machine, using a locally generated dataset with posts obtained from the Twitter API. Through a methodological approach that includes data cleaning, feature selection, and parameter optimization, the Naive Bayes model achieved an accuracy of 95.40%, outperforming Support Vector Machine with an accuracy of 90.45%. The research demonstrates that the quality and representativeness of the dataset are critical to the performance of the models. Additionally, this contribution not only offers an innovative approach adapted to the local context but also opens possibilities for future applications in sentiment analysis and natural language processing in Peru.