The integration of ChatGPT into educational systems underscores the critical importance of reevaluating the relevance of digital literacy skills. This study empirically tested, for the first time, the theoretical premise that the use of ChatGPT could render certain digital skills obsolete by assuming competencies that students were previously required to master. The study aimed to propose an instructional approach that combines a problem-based learning with ChatGPT activity (AIPBL). Subsequently, it sought to quantitatively measure the impact of AIPBL on students' digital literacy competencies, comparing it to a similar PBL assignment devoid of AI technology (PBL) and a control group engaged in lecture-based learning (LBL). In addition, students' experiences with ChatGPT were qualitatively evaluated. The study enrolled 223 healthcare management students, collecting 441 observations through pre-post-tests. Analysis of covariance and thematic analysis of qualitative data were employed. The results revealed that AIPBL led students to utilize fewer digital literacy skills compared to the PBL group, which displayed improvement across all three digital literacy components. Specifically, AIPBL did not significantly enhance participants' perceptions of access and analysis skills, but it did bolster their evaluation skills between pre- and post-tests. Conversely, the LBL group showed no significant differences between their pre- and post-test results. In conjunction with qualitative findings, this study contributes to existing research by affirming the necessity for updated AI literacy skills in education and highlighting the evolving role of literacy skills in the context of AI integration in higher education.