Estimating the impact of Industry 4.0 automation on curricular competence indicators in Brazilian vocational education and training: A mixed-methods AI-supported analysis

Context: The Fourth Industrial Revolution has accelerated the integration of automation technologies into the world of work, raising important questions about the future of Vocational Education and Training (VET). While existing literature has primarily focused on the labor market impacts of automation, few studies have investigated its direct effects on VET curricula. This article addresses this gap by assessing how automation may influence the structure and content of technical courses offered by Brazil's National Service for Commercial Apprenticeship (Senac), one of the country's largest VET providers. Approach: We implemented a three-stage methodology to estimate the impact of automation on technical education: (i) Technological mapping, (ii) prompt development, and (iii) assessment. In the third stage, we combined human expertise with generative Artificial Intelligence tools (GPT-4 and Claude 2) to evaluate 2,100 Course Competency Indicators (CCIs) across 35 technical courses. This dual approach enabled a scalable yet context-sensitive analysis, leveraging both the depth of human judgment and the efficiency of AI. Findings: The technological mapping identified seven key categories of automation technologies: 3D/4D Printing and Modeling, Applied AI, Data Analytics, Digital Platforms and Applications, Extended Reality, IoT and Connected Devices, and Robotics. The developed prompt provided structured guidance for assessing automation impact on CCIs, including instructions for classifying technologies, estimating impact levels, and justifying the results. The assessment showed that 70.3% of the CCIs are at Medium (39.1%) or Low (31.2%) levels of automation impact, suggesting that the courses remain current and relevant, challenging the narrative of rapid obsolescence in technical education. Digital Platforms and Applications were the most frequently cited technology, appearing nearly three times more often than Applied AI and Data Analytics. In contrast, 3D/4D Modeling and Extended Reality had limited relevance in the current course content. Conclusions: This research contributes to global discussions on the future of VET in the context of rapid technological change. It also highlights how automation risk assessments can support curriculum development by identifying where updates or innovations are most needed. Strengthening the alignment between training programs and emerging labor market demands will be essential to ensuring inclusive, future-oriented vocational education.