This article addresses the current problem of hash table security in the context of DoS attacks caused by deliberate collisions. Hash collisions can be exploited by attackers to significantly slow down system performance, as this increases the time for searching and inserting data in hash tables. The main focus is on protective mechanisms implemented in Python, where randomized salts are used in hash functions to prevent such attacks. The paper provides a detailed analysis of how hash tables work, the nature of collision occurrences, and the features of implementing randomized hashing. Key concepts are explained, and practical examples in Python are provided. Various collision attack methods and their prevention strategies are explored. The concluding part of the article contains specific recommendations for the secure use of hash tables and enhancing the overall security of information systems. Both the technical aspects of implementing protective mechanisms and the organizational measures for protection against DoS attacks are considered. The article will be useful for software developers, information security specialists, and researchers in the field of computer science.