As the digital landscape continues to expand, so, too, does the need for regulatory mechanisms – not least in the realm of artificial intelligence (AI). Across the globe, a key question has emerged: how can auditors benefit from the power of AI in a way that is ethical and responsible?
This question was at the top of the agenda at the recent Supreme Audit Institutions 20 (SAI20) summit held in Goa, India. An offshoot of the G20 forum, the SAI20 group was established in recognition of the role of SAIs in promoting the efficiency, accountability and transparency of public administration, and in supporting the implementation of the UN’s sustainable development goals.
At the summit, Girish Chandra Murmu, India’s Comptroller and Auditor General (CAG), highlighted the dual issues of the need to use AI in audit while also establishing a regulatory structure for auditing systems that use the new technology. ‘With AI making greater inroads into governance, SAIs must inevitably prepare themselves for auditing AI-based governance systems,’ he told the group. ‘Simultaneously, SAIs must look for opportunities to adopt AI into their audit techniques to increase their effectiveness.’
AI systems may generate flawed outcomes that can have far-reaching consequences
Critical concerns
A key concern is data security. As AI systems process vast amounts of data, safeguarding this information becomes paramount to maintaining the integrity and trustworthiness of the outcomes generated.
Another pressing challenge lies in the potential for inadvertently perpetuating disparities in the training data or algorithmic design. This can lead to systemic biases and unequal treatment, further amplifying existing societal inequalities.
There is also a risk that when improperly designed, trained or implemented, AI systems may generate flawed outcomes or decisions that can have far-reaching consequences. This emphasises the need for rigorous testing, validation and continuous monitoring of AI systems to ensure their outputs are fully aligned with ethical standards.
Developing and deploying AI requires prioritising ethical and responsible approaches
During the SAI20 summit, Murmu highlighted that by having AI as an ally, the CAG’s office was able to detect false claims made to the government’s Digital Saksharta Abhiyan programme, which aims to promote digital literacy across the country. Using an automated process that achieved a level of analysis impossible through manual assessment alone, multiple instances of fraud were detected including identifying a single person claiming a benefit more than 100 times.
While the potential for AI-based governance systems and data-driven audits is vast, Murmu stressed the need for responsible AI practices where fairness is upheld, and ethics are the cornerstone. Those values are articulated via INDIAai, a groundbreaking web portal that is a knowledge hub, research organisation, and an ecosystem building initiative. The portal promotes collaborations with entities to grow India’s AI ecosystem while ensuring that national policy on AI enforces consistency, reliability, transparency, and accountability, while safeguarding human values and legal compliance.
The way forward
As auditors strive to balance AI-powered audits with human expertise, they are navigating uncharted territory. This captivating dance between technology and tradition requires auditors to weave their way through intricate algorithms and data, embracing the transformative power of AI.
Developing and deploying AI requires prioritising ethical and responsible approaches. This includes promoting diversity within development teams; conducting thorough audits and assessments of AI models; fostering transparency in systems; and implementing robust governance frameworks that uphold fairness, accountability and human rights.
By proactively addressing these concerns, auditors can harness the transformative potential of AI while helping to build a trustworthy and inclusive digital future.
