AUTHOR: Monique Verduyn
A recent SAICA webinar explored the potential of AI in auditing, demonstrated practical applications, and addressed the critical questions that will shape the future of the profession.
‘Artificial intelligence is one of the most profound things we’re working on as humanity. It is more profound than fire or electricity,’ Sundar Pichai, CEO of Alphabet, said in 2020.
And driven by the rapid advancements in AI, the audit profession is changing. A recent SAICA webinar, ‘Audit Reform: Digital Transitioning: Digital Tools (AI) and Audit’ hosted by Msizi Gwala, Project Director for Enabling Competencies at SAICA, explored the transformative potential of AI in the audit profession.
As industries worldwide grapple with the complexities of digital transformation, auditors are uniquely positioned to leverage AI technologies to enhance efficiency, accuracy, and overall audit quality. The full adoption of AI in auditing is expected within the next 2−15 years, depending on factors such as regulatory developments and the evolution of AI technologies. As AI tools become more integrated into audit processes, auditors must continuously upskill and adapt to the changing landscape.
‘In this transition period,’ Gwala noted, ‘auditors are being required to stay abreast of technological advancements and their implications for audit practices. AI is a powerful tool for improving audit processes but also a potential source of new risks and ethical dilemmas.’
Tammy Patuel, SA Data Delivery Leader at EY, emphasised that while AI can significantly enhance efficiency, accuracy, and risk detection in audits, the core focus must remain on ensuring audit quality. ‘Effective AI deployment should result in improved accuracy, enhanced risk detection, deeper insights, and automation of routine tasks, but it is crucial to balance these benefits with a firm commitment to audit quality.’
Opportunities with AI
AI introduces several opportunities for the audit profession:
- Improved accuracy and risk detection − AI can analyse vast datasets with speed and precision far exceeding human capabilities. This allows auditors to process and interpret large volumes of data in real time, leading to timely and accurate risk detection.
- Deeper insights − AI’s ability to learn from past data and apply this knowledge to current audits can revolutionise the detection of fraud and financial anomalies. AI can identify patterns, anomalies, and trends that might indicate risks or errors, providing auditors with deeper insights into the audited entities.
- Automation − AI can automate routine tasks such as data gathering, reconciliations, and form completion. This allows auditors to focus on high-level analysis and complex judgement areas, thereby improving overall audit quality and making continuous auditing a tangible reality.
Challenges and risks
The integration of AI into auditing comes with its own set of challenges and risks, the foremost being data privacy and security. Audits involve handling vast amounts of sensitive data, making the protection of this data paramount. Organisations must establish robust protocols to safeguard client information, ensuring that any breaches or leaks are prevented. In addition to data security, the deployment costs of AI tools can be substantial.
‘We need to carefully assess the financial implications of implementing AI technologies and ensure that these investments deliver significant returns in terms of enhanced efficiency and accuracy in auditing processes,’ Patuel said.
Another critical challenge is regulatory compliance. Auditors must navigate a complex regulatory environment to ensure that AI applications adhere to data protection laws and other relevant regulations. This requires close collaboration with regulators to align AI deployments with legal standards. Additionally, there is a significant risk of bias in AI systems, which can arise from biased training data or flawed algorithmic design.
To address this, it’s crucial to develop AI systems with fairness and objectivity in mind, continuously monitoring and mitigating any biases that may emerge. This vigilance helps maintain the integrity and reliability of AI-enhanced auditing processes.
AI in action
Practical applications of AI in auditing demonstrated how AI tools can modernise the audit process by enhancing efficiency, accuracy, and reliability. Patuel explained how AI can gather and evaluate data from various sources, offering benchmarking data and ratios directly in audit files. This allows auditors to make informed decisions based on comprehensive market and peer analysis. For instance, an AI tool can compare a company’s financial ratios with industry standards, helping auditors identify any discrepancies or areas of concern.
AI-driven risk assessment tools can analyse vast datasets to identify patterns and anomalies that might indicate risks. By learning from past data, AI can highlight potential risk areas that may require further investigation, which is crucial for accurately assessing the overall risk profile of the audited entity.
AI can also verify the authenticity of hundreds or thousands of documents, detecting any alterations or tampering, which is crucial for maintaining the integrity of audit evidence. By examining the content and structure of documents, AI can detect signs of fraudulent activities. For example, if an AI system identifies that financial statements have been altered after approval, it can flag this for further investigation.
Centralised data platforms, powered by AI, offer a holistic approach to data management and analysis in audits. These platforms can ingest distinct types of data, including structured data (like financial statements) and unstructured data (like emails and contracts), allowing auditors to view and analyse data from multiple sources in one place. AI-driven anomaly detection can identify irregularities across different audit sections, ensuring a comprehensive audit approach.
Real-world applications and examples
Amé Thwaits, Innovation and Digital Leader for EY Africa Assurance, provided examples of how AI is being implemented in the audit profession. One significant application is in audit planning and risk assessment.
‘AI tools can evaluate benchmarking data and financial ratios during the planning phase, drawing on previous audits and market data to suggest specific risk areas and focal points for the audit team,’ she noted. ‘This capability enables auditors to anticipate potential issues and allocate resources more effectively, enhancing the overall audit strategy.’
AI also plays a crucial role in assisting with audit methodology, Thwaits explained. ‘AI tools embedded in audit files can provide real-time guidance, offering advice and summaries of relevant audit standards. This helps auditors make informed decisions quickly and ensures that their methodologies align with current best practices.’
AI can also automate routine tasks such as data gathering, reconciliations, and form completion. By handling these repetitive activities, it reduces the risk of human error and frees auditors to concentrate on more complex and judgement-intensive aspects of the audit, ultimately improving the quality and efficiency of the audit process.
Enhanced decision-making
Improved decision-making is the auditor’s dream; AI promises to make that a reality by providing deeper insights and supporting continuous auditing.
‘By analysing large datasets to identify trends and patterns, AI can give auditors a comprehensive understanding of the audited entity’s operations and financial health,’ Thwaites said. ‘This ability to extract actionable insights from vast amounts of data will help the profession make more informed decisions and offer better recommendations to their clients.’
In addition, AI supports continuous auditing by providing ongoing assurance and real-time analysis of financial data. This non-stop monitoring allows auditors to detect and address issues as they arise, rather than waiting for periodic audit cycles.
‘Continuous auditing ensures that financial records are always up-to-date and accurate, increasing the reliability and relevance of the audit findings, she added. ‘By leveraging these AI capabilities, auditors can enhance their decision-making processes, improve audit quality, and provide greater value to stakeholders.’
Skills for the future
The speakers agreed that to remain competitive and effective auditors need to develop new skills. Data literacy is crucial; understanding and analysing data are fundamental to leveraging AI effectively. Thwaits noted that members of the profession must become proficient in data management and interpretation to draw meaningful insights from complex datasets. Computational thinking is another essential skill; auditors need to approach problems logically and devise solutions using AI and other technologies. Collaboration is key. Working closely with data scientists can help auditors exploit AI’s full potential. Adaptability and a commitment to continuous learning are vital to keep pace with technological advancements.
Regulatory and ethical considerations
The discussion highlighted that compliance with relevant laws and standards is critical to maintaining the integrity and trustworthiness of audits. Several key points emerged emphasising the need for a robust regulatory framework and adherence to ethical principles.
Working with regulators
Patuel stressed that collaboration with regulators is critical to align AI applications with existing and emerging regulations. Compliance with data protection laws and other regulations is crucial for safeguarding client data and ensuring privacy. Understanding the regulatory landscape and implementing measures to adhere to these laws is a foundational step. Regulators can also help establish standards for AI use in auditing to ensure consistency and reliability across the profession.
Standardisation facilitates the development of best practices and guidelines for implementing these tools. The regulatory environment is dynamic with new laws and amendments being introduced regularly, which is why it is essential to work closely with regulators to stay updated on these changes and adapt accordingly.
Ethical considerations
The discussion highlighted several key ethical principles that must guide the deployment of AI tools. Transparency is crucial, as AI systems should provide clear insights into how decisions and analyses are made. This transparency helps auditors understand and trust AI’s outputs and explain these outputs to stakeholders. Fairness is another important principle; AI tools must be designed and trained to avoid biases that could lead to inaccurate or discriminatory outcomes. This requires using unbiased training data and carefully monitoring AI systems for any signs of bias. Accountability is also essential. Patuel warned that the profession must remain accountable for the use of these tools, ensuring they are used ethically and responsibly and that any errors or biases are promptly addressed.
Establishing a regulatory framework
The speakers discussed several aspects of designing new regulatory frameworks for AI. For example, at EY, an AI assurance framework has been developed to assist with the auditing of AI systems. This framework covers understanding AI strategies, management controls, and risk considerations, which are essential for ensuring that AI systems are reliable and trustworthy.
The concept of explainable AI is also critical for maintaining transparency and accountability. Patuel noted that explainable AI systems provide insights into how decisions are made, helping auditors and stakeholders understand AI’s processes and outputs
‘The regulatory framework will have to evolve alongside advancements in AI technology,’ she added. ‘Because current regulations might not fully address the complexities of AI, ongoing updates and enhancements to the standards will be required.’
Ethical and professional judgement
Professional judgment remains the cornerstone of the auditing profession. AI lacks the ability to understand context, interpret complex situations, and apply ethical considerations in scenarios that are not straightforward or easily quantifiable. Only humans can perform these critical tasks.
While data analytics can provide valuable insights, human judgement is indispensable for making nuanced decisions that consider the broader context and specific circumstances of each audit. This human element ensures the integrity and reliability of the audit process.
AI as a co-pilot
Through AI, the profession is set to gain the benefits of new processes, insights, and capabilities that can unlock true creative potential. As Steve Jobs once said, ‘Technology alone is not enough. It’s technology married with the liberal arts, married with the humanities, that yields us the results that make our hearts sing.’
This article was first published by Accountancy SA September Edition at the following URL: http://magazine.accountancysa.org.za/asa-september-2024?m=52861&i=829822&p=1&ver=html5
