ENGLISH:
This study was aimed at evaluating the effectiveness of Semantris Google AI in improving vocabulary achievement, particularly synonyms, meronyms, and hyponyms of nouns, among 71 seventh grade students at SMPN 1 Mendobarat. These students were divided into an experimental group of 36 students using Semantris and a control group of 35 students using non-Semantris in a quasi-experimental design. Data were collected through observation sheets and vocabulary tests. Pre- and post-tests in the form of vocabulary assessments were administered to both groups, and data were analysed using paired sample t-tests and independent sample t-tests. The results showed dynamic vocabulary progress in both groups during the intervention. The paired sample t-test confirmed a significant increase within each group, with Sig. 2-tailed 0.00 less than 0.05 for both Semantris and non-Semantris. However, the independent sample t-test showed Sig. 2-tailed 0.255 greater than 0.05, rejecting the alternative hypothesis and indicating no significant difference between the groups. Therefore, Semantris could effectively improve vocabulary achievement but was not superior to the non-Semantris approach. This highlighted the potential synergy between artificial intelligence and human-guided instruction for adaptive and inclusive teaching in lower secondary school English classes