One of the important technical requirements of construction materials is strength. Construction materials before being used need to be tested and evaluated for strength. Current standards generally prescribe the determination or prediction of the strength of materials through destructive, non-destructive testing methods or Ihrough experimentally established theoretical relationships. These methods are often costly and time consuming, but often have low accuracy due to 1he limited number of experiments or the difficulty of fully establishing the factors affecting the strength of the material. Artificial intelligence (Al) is a new solution with a focus on machine learning (ML) algorithms that can build predictive models of the properties of building materials, including strength, with high accuracy, fast and cost saving. This paper introduces the machine learning methods being used effectively and the principles of using those methods to predict the strength of construction materials. Recommendations on solutions for effective application of the ML methods are also mentioned so that artificial intelligence becomes an effective method of predicting the strength of construction materials in Vietnam.