The integration of artificial intelligence (AI) in L2 teaching and learning is poised to revolutionise educational practices by enhancing both instructional methods and language development for L2 learners. This study employed a mixed-methods design to comprehensively examine the impact of AI tools, machine translation systems, and traditional approaches on students' translation accuracy, emotions, and motivation. A total of forty-nine undergraduate English majors were divided into three groups: the AI Group (AIG; N = 16) using AI tools, the machine translation group (MTG; N = 20) using machine translation tools, and the traditional group (TG; N = 13) using manual methods. Participants completed four translation tasks with varying levels of linguistic complexity, and their performance was evaluated using quantitative metrics such as meaning retention, grammatical correctness, fluency, and naturalness. Additionally, semi-structured interviews were conducted to gather qualitative insights into participants' emotional and motivational experiences. Quantitative data analysis included the Kruskal-Wallis test to assess differences amongst the groups, revealing that AIG students achieved the highest translation accuracy. Qualitative thematic analysis of the interview data indicated that emotions such as curiosity, anxiety, and excitement were prevalent across all groups. While AI tools fostered motivation in the AIG and MTG, some participants expressed concerns about over-reliance on technology leading to reduced engagement. These findings highlight AI's dual role in enhancing translation accuracy and shaping the emotional and motivational dynamics of L2 learners, suggesting that its integration should be balanced with traditional methods to optimise learning outcomes.