Abstract
This article examines the effectiveness of Artificial Intelligence tools in assessing English speaking and writing skills in contemporary language education. The study explores the theoretical foundations of AI-based assessment, with particular attention to automated writing evaluation and speech recognition technologies. It reviews existing digital platforms used for evaluating productive skills and analyzes their reliability, validity and alignment with CEFR standards. In addition to the theoretical overview, the article presents the results of an experimental study conducted with language learners at intermediate levels. The experiment investigates how AI-generated feedback compares to teacher assessment and evaluates the consistency and accuracy of automated scoring. The findings reveal significant advantages of AI tools, including instant feedback, scalability and scoring consistency while also identifying important limitations in assessing higher-order skills such as coherence, argumentation and communicative effectiveness. The study is concluded by discussing pedagogical implications and recommending a balanced, hybrid assessment approach that combines AI technologies with human expertise.