Agentic AI Microservice Framework for Deepfake and Document Fraud Detection in KYC Pipelines

Authors: Chandra Sekhar Kubam

Published: 2026-01-09 17:01:40+00:00

Comment: Journal of Information Systems Engineering and Management, 2024

AI Summary

This paper introduces an Agentic AI Microservice Framework to enhance deepfake and document fraud detection within Know Your Customer (KYC) workflows, addressing the limitations of traditional monolithic systems. The framework integrates modular vision models, liveness assessment, deepfake detection, OCR-based document forensics, and multimodal identity linking, orchestrated by autonomous micro-agents. Experimental evaluations demonstrate improved detection accuracy, reduced latency, and enhanced resilience against adversarial inputs for robust, real-time KYC verification.

Abstract

The rapid proliferation of synthetic media, presentation attacks, and document forgeries has created significant vulnerabilities in Know Your Customer (KYC) workflows across financial services, telecommunications, and digital-identity ecosystems. Traditional monolithic KYC systems lack the scalability and agility required to counter adaptive fraud. This paper proposes an Agentic AI Microservice Framework that integrates modular vision models, liveness assessment, deepfake detection, OCR-based document forensics, multimodal identity linking, and a policy driven risk engine. The system leverages autonomous micro-agents for task decomposition, pipeline orchestration, dynamic retries, and human-in-the-loop escalation. Experimental evaluations demonstrate improved detection accuracy, reduced latency, and enhanced resilience against adversarial inputs. The framework offers a scalable blueprint for regulated industries seeking robust, real-time, and privacy-preserving KYC verification.

Key findings

The framework achieved a deepfake detection recall of 91.3% with temporal liveness and artifact detection, and 93.1% with transformer-based multimodal models, outperforming a CNN baseline by up to 20% recall. Document fraud detection reached 96.1% accuracy for distinguishing authentic from synthetic documents. End-to-end KYC verification averaged 2.7 seconds, with agentic orchestration reducing microservice failures by 35% and improving anomaly recall by 15%.

Approach

The proposed framework utilizes an Agentic AI Microservice Architecture. It integrates specialized microservices for tasks like vision liveness & deepfake detection, document OCR & template verification, and identity linking. These services are orchestrated by autonomous micro-agents that handle task decomposition, dynamic model selection, failure recovery, anomaly escalation, and policy compliance, ensuring a scalable and resilient workflow with human-in-the-loop escalation for complex cases.

Datasets

A selfie dataset (20,000 genuine, 10,000 spoofing attempts, 5,000 AI-generated deepfake videos) and a document dataset (50,000 authentic ID cards, 10,000 synthetic or forged documents generated using GANs and diffusion-based models).

Model(s)

CNN-based detectors, Temporal Liveness cues, artifact detection, Transformer-based Multimodal architectures, OCR-based text extraction, and template deviation models.

Author countries

USA

← Previous