Background: Artificial intelligence (AI) tools have been increasingly utilised in the production of educational media, including animated educational videos (AEVs) incorporating pedagogical agents (PAs). These tools support the efficient creation of multimedia content and reduce teachers' technical workload. Objectives: This study investigates the relationships among factors influencing teachers' acceptance of AI-assisted tools for creating AEVs, based on the technology acceptance model (TAM). Methods: A mixed-methods approach was employed, combining quantitative analysis using the AI Technology Acceptance Scale (AITAS) based on the TAM and qualitative insights from thematic analysis of open-ended responses from 1569 teachers. Structural equation modelling was applied for the quantitative analysis, confirmed an excellent model fit (CFI = 0.97, GFI = 0.95, RMSEA = 0.06, SRMR = 0.03). Sentiment analysis further supported the qualitative findings. Results and Conclusions: The results show that perceived usefulness and ease of use significantly influence teachers' attitudes towards adopting AI tools for AEV production. These tools improve content quality and student engagement. While 77.44% of teachers expressed positive attitudes toward using AI, technical challenges were also reported, particularly in relation to limited skills, inadequate hardware, and poor internet connectivity. These issues were more prevalent in less-resourced areas. To support successful integration, it is essential to provide targeted training and ensure broader access to AI tools.