Comparative Evaluation of AI-Generated Text Detection Approaches

this study presents a comparative evaluation of three approaches for AI-generated text detection. Experimental results demonstrate that the RoBERTa model significantly outperforms all other models. The study concludes that deep contextual embeddings offer the most reliable solution for modern AI-text detection, while classical approaches remain effective for computationally constrained environments.