Voice recognition technology has made significant strides in many languages, yet the development of accurate and efficient Arabic voice recognition systems remains a challenge. For Automatic Speech Recognition (ASR) systems, Arabic with its various dialects, a rich structure, and pronunciation variations, is a challenge. In this research, Google Translation AI integration is studied with the current voice recognition technique to enhance the accuracy of Arabic voice recognition by concentrating on the traditional Iraqi slang language. For the proposed system, the dataset of 20 traditional Iraqi words was tested and an overall accuracy of 96.2% was achieved. Results show that the application of AI based translation with traditional ASR techniques can benefit the voice recognition systems for Arabic and other linguistically complex languages.