The application of intelligent technologies in the field of pavement material testing has improved test efficiency and data accuracy, yet it has also given rise to new types of research integrity risks. Taking the postgraduate course Pavement Material Testing Technology as the research object, this paper analyzes typical integrity risks, including black-box algorithms and selective data reporting in pavement material testing. Corresponding solutions are designed and implemented in teaching activities: embedding ethical modules into theoretical lectures, integrating full-process integrity training into experimental operations, incorporating team collaboration norms and intellectual property requirements into classroom discussions, and adding academic standards to course paper assessments. Follow-up surveys show that this collaborative education practice has effectively improved postgraduate students’ ability to identify technical risks, standardize their collaborative behaviors, and enhance their awareness of intellectual property rights. This study provides practical references for transforming research integrity education in postgraduate courses from explicit preaching to implicit integration.