Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{196092,
        author = {K.Janvitha and B.Srithika and E.Jyoshna and D.Nivedhitha},
        title = {Agentic AI Auditor},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {1487-1493},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196092},
        abstract = {The rapid proliferation of Artificial Intelligence (AI) systems across enterprise, healthcare, finance, and government sectors has created an urgent need for rigorous, automated auditing and compliance verification frameworks. Traditional manual auditing processes are inadequate to keep pace with the speed, scale, and complexity of modern AI deployments. This paper proposes an Agentic AI Auditor — an autonomous, multi-agent system capable of continuously monitoring AI pipelines, verifying regulatory compliance, detecting anomalous model behavior, and generating structured audit reports without human intervention.
The proposed system leverages agentic AI principles, wherein autonomous agents equipped with planning, reasoning, and tool-use capabilities work collaboratively to audit diverse AI systems. The framework encompasses four primary audit dimensions: model fairness and bias detection, data integrity and lineage verification, regulatory compliance mapping (including GDPR, EU AI Act, and ISO/IEC 42001), and runtime behavioral anomaly detection. Each dimension is handled by specialized sub-agents coordinated by a central orchestrator agent.
The system integrates a secure role-based access control mechanism, immutable audit trail logging using cryptographic hashing, and a real-time compliance dashboard. A structured pipeline covering data ingestion, feature analysis, policy rule evaluation, and report synthesis is implemented to maximize audit coverage and accuracy.
Experimental evaluation across multiple AI system types demonstrates that the proposed Agentic AI Auditor achieves high compliance detection rates while minimizing false compliance certifications.
By combining autonomous agent reasoning, machine learning-based anomaly detection, and regulatory rule engines, the proposed solution effectively enhances accountability, transparency, and trustworthiness in AI deployments.},
        keywords = {Agentic AI, AI Auditing, Compliance Verification, Multi-Agent Systems, Model Fairness, Regulatory AI, Anomaly Detection, AI Governance.},
        month = {April},
        }