The operation time of a disposal repository is generally more than one hundred years except for the institutional control phase. The structural integrity of a repository can be regarded as one of the most important research issues from the perspective of a long-term performance assessment, which is closely related to the public acceptance with regard to the nuclear safety. The objective of this study is to suggest the methodology for quantitative evaluation of structural integrity in a nuclear waste repository based on the adaptive artificial intelligence (AI), fractal theory, and acoustic emission (AE) monitoring. Here, adaptive AI means that the advanced AI model trained additionally based on the expert’s decision, engineering & field scale tests, numerical studies etc. in addition to the lab. test. In the process of a methodology development, AE source location, wave attenuation, the maximum AE energy and crack type classification were subsequently studied from the various lab. tests and Mazars damage model. The developed methodology for structural integrity was also applied to engineering scale concrete block (1.3 m × 1.3 m × 1.3 m) by artificial crack generation using a plate jacking method (up to 30 MPa) in KURT (KAERI Underground Research Tunnel). The concrete recipe used in engineering scale test was same as that of Gyeongju low & intermediate level waste repository. From this study, the reliability for AE crack source location, crack type classification, and damage assessment increased and all the processes for the technology development were verified from the Korea Testing Laboratory (KTL) in 2022.