• Version
• Download 7
• File Size 489.14 KB
• File Count 1
• Create Date 22 March 2026
• Last Updated 23 March 2026

AI Based Synthetic Identity Detection in KYC Systems

A. Ramya Sri
Pranavi Damanapally
Department of Computer Science Jyothishmathi
Institute of Technology and Science
Karimnagar, Telangana, India
pranavidamanapally@gmail.com

Sree Nikitha
Department of Information Technology
Jyothishmathi Institute of Technology and Science
Karimnagar, Telangana, India
22.6a9sreenikitha@gmail.com

K. Saakshith
Department of Computer Science Jyothishmathi
Institute of Technology and Science
Karimnagar, Telangana, India
kattasaakshith@gmail.com
A. Aravind
Department of Information Technology Jyothishmathi
Institute of Technology and Science Karimnagar,
Telangana, India aravindchitumalla4@gmail.com
V. Veda Sree
Department of Information Technology Jyothishmathi
Institute of Technology and Science Karimnagar,
Telangana, India vedasree@gmail.com

Abstract—The rapid growth of digital banking and online financial services has created a strong need for secure and reliable identity verification systems. Traditional Know YourCustomer (KYC) processes mainly rely on manual document  checks and physical verification, which can be slow, costly, and sometimes prone to human errors. With the increasing number of digital fraud cases and identity theft, financial institutions require smarter solutions to ensure the authenticity of user identities during the onboarding process.This project proposes an AI-based KYC fraud detection system that aims to automate and strengthen the identity verification process. The systemuses Optical Character Recognition (OCR) to extract important information from uploaded identity documents such as name, identification number, and date of birth. It also uses facial recognition technology to compare the photo present on the document with a live selfie provided by the user. By combiningdocument verification and biometric authentication, the system  improves the accuracy of identity validation.Additionally, thesystem analyzes the extracted data and biometric results to  identify possible inconsistencies that may indicate fraudulent or synthetic identities. The proposed approach helps reduce manual effort, speeds up the verification process, and improvesthe reliability of digital KYC systems.Index Terms—Synthetic Identity Fraud, Know Your Customer,OCR, Facial Recognition, Machine Learning, Fraud Detection, Digital Onboarding