This entry critically outlines how artificial intelligence (AI) has prompted a fundamental rethinking of language assessment in computer-assisted language learning (CALL), and raises new questions about validity, reliability, and equity. Tracing developments from early digitized tests to contemporary applications of large language models (LLMs), it highlights how AI technologies now mediate task generation, learner interaction, feedback delivery, and scoring. While these tools claim to offer greater efficiency and personalization, they also challenge core assumptions about language assessment. The entry argues that persistent issues surrounding bias, data privacy, and accountability highlight the need for ethical and pedagogically grounded frameworks. Finally, the integration of generative AI encourages educators, researchers, and policymakers to reconsider not only how language is assessed but also why it is assessed.