<![CDATA[The integration of artificial intelligence in science learning is a strategic need to address the challenges of scientific literacy and sustainable education. This study aims to develop and test the effectiveness of a STEAM-ESD (Science, Technology, Engineering, Arts, and Mathematics-Education for Sustainable Development)-based chatbot in improving students' scientific literacy in biotechnology learning. The study used a research and development (R&D) design with the ADDIE model which includes the stages of analysis, design, development, implementation, and evaluation. The product developed is an AI chatbot "BIOTEKS" along with learning support tools. The research subjects were 12th grade high school students. Data were collected through expert validation sheets and science literacy tests, then analyzed descriptively and inferentially using Rasch analysis, N-gain, and the Mann-Whitney U test. The results showed that all learning tools had a very high level of validity and reliability (94–97%), making them feasible to implement. The implementation of the STEAM–ESD-based BIOTEKS chatbot significantly improved students' scientific literacy with high N-gain values in both classes (0.99 and 0.77), as well as significant Mann-Whitney U Test values between classes (p = 0.199). The most prominent improvement occurred in scientific representation and data interpretation abilities in the context of sustainable biotechnology. It can be concluded that the STEAM–ESD-based BIOTEKS chatbot is effective as an innovative learning medium to improve students' scientific literacy and serves as a pedagogical reinforcer in contextual and sustainability-oriented 21st-century biology learning.]]>
