This study aims to analyze students’ self-efficacy and science process skills as a basis for developing a Problem-Based Learning (PBL)-based physics laboratory design integrated with Artificial Intelligence (AI). A mixed-methods approach was used with a sample of 46 high school students. Data were collected via a questionnaire and analyzed using quantitative descriptive methods based on percentage categories. The results of the study indicate that students’ self-efficacy was in the range of ±70% (high category: 51%–75%), science process skills at ±69% (high category: 51%–75%), and perceptions regarding the use of AI at ±68% (high category: 51%–75%). Some indicators even reached the very high category (76%–100%), such as the use of AI in understanding physics concepts (80.43%). However, some indicators were still in the low category (26%–50%), such as independence in designing experiments (46.74%) and doubts and concerns regarding AI use (45.65%–49.46%). Additionally, there was a tendency for students to accept information from AI without critical verification. These findings highlight the need to develop PBL-AI-based physics laboratory designs that not only integrate technology but also emphasize strengthening students’ independence, critical thinking skills, and the validation of scientific information.
Copyrights © 2026