This study explored the way Saudi EFL students speak English after using speech recognition software operated by artificial intelligence. To obtain data about students' experience and observable improvement, researchers used questionnaires and pre- and post-tests to assess improvement. During the span of eight weeks, 45 Saudi intermediate students worked with Speechling, a computer-based voice recognition system. Speechling gave students immediate feedback about their pronunciation of English and a structured approach for practicing pronunciation. In order to get accurate information about students' attitudes towards the technology and how they responded to it, researchers administered them surveys along with spoken tests. The tests measured how much pronunciation and fluency had improved and the questionnaires asked about feelings towards. In addition to testing by speech, researchers gave students questionnaires to get detailed information on what they felt about the technology and how they responded to it. The tests measured how much pronunciation and fluency had improved and the surveys on attitudes on and feelings towards the technology. The participants' fluency and confidence in English speech were improved after going through AI-SRT. Their average pronunciation score was much better from 7.2 to 14.6 (Cohen's d = 2.4). The same outcome was observed in the fluency tests: participants' rhythm was improved and fluency was higher with reduced hesitations. They sounded more natural overall. Most of the participants had a positive attitude towards using AI-SRT. 78% of the respondents liked getting instant feedback and almost 82% reported it had helped their standard classroom learning. Understanding of the connection between language and AI learning is made possible through this initiative. A perfect example of how this approach of teaching improves speaking ability and pronunciation is the utilization of speech recognition software by English language learners of Arabic origin.