From the analysis of recent researches of automatic question generation using deep learning techniques, we examined papers between 2022 and early 2023 from the examination of recent research on automatic question production using deep learning techniques. Our study comes after the survey report that broadens the analysis of earlier evaluations of AQG content that surfaced between late 2014 and early 2019. We examined the researched works that were included, looking at things like the (1) framework for question generation and (2) generating method. We discovered that contemporary methods fre- quently rely on generative frameworks that deploy Transformer-based models and GPT-n series, which are more efficient in terms of analysis and perfor- mances. We discovered that question creation has gained popularity recently and has significantly improved the educational field. Yet, it can be challeng- ing to produce automatic questions and create the necessary question patterns, structures, and forms. Our additional research advises testing out more prac- tical, efficient models and strategies for autonomous question generating.