Age-related macular degeneration (AMD) is a progressive degenerative retinal disease and a leading cause of blindness in older adults worldwide. According to numerous studies, the number of affected individuals reached 196 million in 2020, with projections estimating an increase to 288 million by 2040, including 18.6 million cases of advanced AMD. The advent of optical coherence tomography (OCT) has enabled researchers and clinicians to characterize microstructural changes in different retinal layers at earlier disease stages and improve monitoring strategies. Important steps have been taken to develop algorithms capable of recognizing early signs of AMD, assessing its severity, and predicting progression. These algorithms have formed the basis for artificial intelligence (AI)-driven systems applicable to any hardware or software exhibiting intelligent behavior. OCT imaging allows for the identification of biomarkers whose presence or interaction with other factors predict transition from intermediate to advanced AMD. The obtained data can provide deeper insights into the pathogenesis of intermediate AMD, enhance early diagnosis for timely intervention, and facilitate the search for new treatment options. Artificial intelligence could make this process easier, simpler, less time-consuming, and more accurate by integrating structural OCT data with genetic risk indicators and lifestyle characteristics. However, the results are still inconsistent due to factors leading to limited result reliability, such as database quality, sample sizes, and data acquisition methods.