<p>In today’s digital landscape, the widespread adoption of Large Language Models (LLMs) has made it increasingly difficult to distinguish between human-written and AI-generated text. This challenge stems from the ability of state-of-the-art LLMs such as Gemini, the GPT series (including ChatGPT-4o, 4.1, and 4.5), LLaMA, and DeepSeek to produce highly sophisticated, human-like text. The indistinguishability of such text poses significant risks across multiple sectors, including cybersecurity threats, propaganda dissemination, the spread of biased or false information on social media, and the facilitation of social engineering attacks. In education, the misuse of these models increased academic dishonesty. To address these challenges, we leveraged two diverse datasets, HC3- and M4GT English, comprising both human-authored and AI-generated text. We evaluated model performance using standard metrics such as accuracy, precision, recall, and F1-score. Our findings demonstrated that the DistilBERT model achieved superior results, with an accuracy of 99.45% on the HC3 English dataset with linguistic features and 96.23% on the M4GT English dataset with linguistic features in distinguishing AI-generated text from human writing. This research contributes to ongoing advancements in AI detection, highlighting the potential of machine learning to address the evolving challenges of content creation and verification.</p>