<![CDATA[The rapid development of artificial intelligence (AI) in the digital learning era has driven significant transformations in science education research, particularly regarding how learning data is analyzed, how concept diagnosis is conducted, and how learning experiences are personalized. This study focuses on mapping the implementation of AI in science education research based on 25 relevant indexed international studies. Using a systematic review approach, articles were analyzed to identify trends, methodologies, contexts of use, and the impact of AI on science learning processes and outcomes. The study results indicate that AI technology in science education is most commonly applied in inquiry-based learning, adaptive learning systems, intelligent tutoring systems, learning analytics, physics misconception detection, automated feedback, augmented reality-based learning, and personalized learning experiences. AI has been shown to improve scientific thinking skills, the accuracy of misconception diagnosis, learning motivation, and the effectiveness of virtual experiments. Meta-findings from 25 sources indicate that the application of AI significantly contributes to improving students' conceptual understanding, scientific argumentation skills, and scientific inquiry performance. However, key challenges identified include teacher readiness, students' AI literacy, infrastructure availability, and the ethical use of learning data. This study concludes that integrating AI into science education research has significant potential to enhance learning quality, but requires sound policy, educator support, and instructional design for optimal and sustainable implementation. ABSTRAK Perkembangan pesat kecerdasan buatan (AI) dalam era pembelajaran digital telah mendorong transformasi signifikan dalam penelitian pendidikan sains, terutama terkait bagaimana data pembelajaran dianalisis, bagaimana diagnosis konsep dilakukan, serta bagaimana pengalaman belajar dipersonalisasi. Kajian ini berfokus pada pemetaan implementasi AI dalam penelitian pendidikan sains berdasarkan 25 studi internasional terindeks yang relevan. Melalui pendekatan kajian sistematis, artikel?artikel dianalisis untuk mengidentifikasi tren, metodologi, konteks penggunaan, serta dampak AI terhadap proses dan hasil belajar sains. Hasil kajian menunjukkan bahwa teknologi AI dalam pendidikan sains paling banyak diterapkan pada pembelajaran berbasis inkuiri, sistem pembelajaran adaptif, intelligent tutoring systems, analitik pembelajaran, deteksi miskonsepsi fisika, umpan balik otomatis, pembelajaran berbasis augmented reality, serta personalisasi pengalaman belajar. AI terbukti meningkatkan keterampilan berpikir ilmiah, akurasi diagnosis miskonsepsi, motivasi belajar, serta efektivitas eksperimen virtual. Meta-temuan dari 25 sumber menunjukkan bahwa penerapan AI memberikan kontribusi signifikan dalam meningkatkan pemahaman konsep, kemampuan argumentasi ilmiah, dan performa penyelidikan ilmiah siswa. Namun, tantangan utama yang ditemukan meliputi kesiapan guru, literasi AI siswa, ketersediaan infrastruktur, serta etika penggunaan data pembelajaran. Kajian ini menyimpulkan bahwa integrasi AI dalam penelitian pendidikan sains sangat potensial memperkuat kualitas pembelajaran, tetapi memerlukan kebijakan, dukungan pendidik, serta desain pembelajaran yang matang agar implementasinya optimal dan berkelanjutan.]]>
