<p>The traditional voting system is a time consuming process and in some cases it will lead to human errors. To avoid this issues, an emerged technology has used in the election process. During the e-voting process, the confidential data of the voters are processed in the internet. However, there are lot of possibilities to steal the user’s data due to less security. To overcome these issues the proposed method used AI based enhanced Twofish encryption with SnS and GCE for secure e-voting. The XGBoost (extreme Gradient Boosting) algorithm is employed to detect the intrusion when the voters access the publicly available link to cast their votes. Based on Scapy, the data are generated from Ethernet, IP, TCP and UDP. Scapy is employs to capture the packets, copy or decode packets, match requests and replies and send packets on the wire. The network discovery, scanning, attacks, trace routing, unit tests and probing are the tasks handled by Scapy. Then generated data undergoes a pre-processing process to remove missing values and convert them into null values. The proposed research uses K-means algorithm for label creation for each data. Once the registration is completed the key scheduling is used to encrypt and verify the user’s data in block chain based environment using enhanced Twofish with SnS (Split n Swap) and GCE (Cyclic Group Extended). The SnS and GCE are used to swap the data order to confuse the attackers and improve the security. After the decryption the voter identification gets verified and they can cast their votes. The proposed model verifies that data in the voter_id have any error or malfunction then the vote will not be submitted. With the help of block chain the data of the voters remains safe and secure. The performance of the proposed model is evaluated by the metrics and attain best performance than existing studies.</p>