Data encryption is being widely employed to secure data in order to provide confidentiality, authenticity and data privacy. In proposed method a novel method for (encryption, decryption) message through optimization using genetic algorithm (GA) is presented which aims to maintain the security of the message via increasing the complexity of the key used in the encryption. The proposed method is named stream cipher randomization using GA (SCRGA). It is characterized by its secrecy by hiding the statistical properties for the input message. The GA increases the key diffusion in order to eliminate the likelihood of using the statistical analysis and cryptanalysis techniques to obtain the original text. A key (k) will be supplied as input to pseudorandom bit generator and then it produces a random 8-bit output, The SCRGA evaluates the candidate keys and it would select the one that achieves the best value of the fitness function by maximization three criteria considered in matching the input message and the key. The test results were based on the comparison of the proposed method with two other state-of-the-art methods, that the SCRGA outperformed the other two methods in terms of the execution time (encryption, decryption) and encryption rounds key size.