Auditing Generative AI

View Sample PDF

Author(s): Asmita Jain (University at Buffalo, SUNY, USA), Zayan Shaheen (University at Buffalo, SUNY, USA), Dollie Juneja (University at Buffalo, SUNY, USA)and Anuja Ulhe (University at Buffalo, SUNY, USA)
Copyright: 2026
Pages: 30
Source title: Advancing IT Audits Through Integrative Approaches and Emerging Technologies
Source Author(s)/Editor(s): Manish Gupta (University at Buffalo, SUNY, USA), John Walp (University at Buffalo, SUNY, USA)and Raj Sharman (University at Buffalo, SUNY, USA)
DOI: 10.4018/979-8-3373-3078-5.ch004

Keywords: Engineering Science Reference / IT Governance / IT Policy and Standardization / Library & Information Science

Purchase

View Auditing Generative AI on the publisher's website for pricing and purchasing information.

Abstract

The swift growth and implementation of Generative AI systems has created several unique governance and compliance concerns, notably in terms of ethical usage, data protection, accountability, and transparency. This study presents a complete audit approach for monitoring GenAI systems, focusing on critical aspects such as risk assessment, control deployment, and management of compliance. By utilizing known governance frameworks such as COSO & COBIT 5 to provide a systematic guideline for examining the operational integrity, security, & ethical aspects of GenAI installations. The approach for the audit framework we propose focuses on addressing GenAI specific risks such as inaccuracies in AI results, data security vulnerabilities, and a lack of clarity in AI decision-making processes. Our method handles these issues using a mix of automated control evaluations, threat evaluation, and continuous monitoring. The study presents ways for resolving governance challenges, such as changing legal environments and the complexities of keeping AI models in accordance with retention laws.

The IRMA Community

Research IRM

Auditing Generative AI

Purchase

Abstract

Related Content

IRMA Sponsors