AI-Assisted science lesson planning for innovative pedagogy in resourceconstrained settings: A scoping review

Abstract

Science teachers, particularly those working in resource-constrained settings, face 
persistent challenges related to limited laboratory facilities, large class sizes, inadequate 
instructional materials, and the substantial time required for lesson preparation and 
resource development. The emergence of generative artificial intelligence (AI), including 
large language models such as ChatGPT, Gemini, Claude, and Copilot has created new 
opportunities to support science teachers in preparing lessons and developing instructional 
resources. However, evidence on how AI is being used for science lesson planning, its      
reported educational benefits, and its associated challenges remains fragmented. This 
scoping review examined the applications, reported benefits, and challenges of AI-assisted 
science lesson planning, with particular attention to its potential contribution to innovative 
pedagogy in resource-constrained settings. Following the existing framework, as advanced 
by previous researchers in about 2 decades ago and reported in accordance with PRISMA
ScR guidelines, a systematic search of peer-reviewed and grey literature published between 
January 2020 and May 2026 identified 40 relevant studies. The synthesis revealed four 
primary thematic categories of AI application in science lesson planning: (1) generating 
lesson materials and content, (2) facilitating virtual laboratory experiences, (3) automating 
planning and documentation, and (4) supporting pedagogical decision-making. Across the 
reviewed literature, reported benefits included reduced lesson-preparation time, improved 
access to curriculum-aligned instructional resources, and support for inquiry-based and 
differentiated teaching approaches. Reported challenges included inaccurate or fabricated 
AI-generated information, insufficient contextualization of instructional materials, unequal 
access arising from digital and infrastructural constraints, and variation in teachers’ capacity 
to critically evaluate AI-generated outputs. The findings suggest that AI can support science 
lesson planning and contribute to innovative pedagogical practices when used as a 
complementary tool rather than a substitute for teachers’ professional judgment. The 
review indicates that effective and responsible implementation requires reliable digital 
infrastructure, critical verification of AI-generated content, contextual adaptation of AI 
outputs, and continuing teacher professional development. These findings highlight both 
the potential of AI to address resource constraints in science education and the need for 
context-sensitive approaches to its integration into lesson planning.