Digital innovation offers unique insights throughout the audit process, enabling auditors to offer increased value and efficiency in a demanding economy.

The second 2023 SAICA Audit Reform Webcast, held in August and chaired by SAICA executive for audit and assurance Thandokuhle Myoli, explored the significant impact of digital technology on the audit profession, both in terms of the tools and techniques used by auditors, and the way in which audits are conducted.

Nevellan Moodley, who leads the BDO Africa fintech desk, guided attendees through various technologies that enhance the efficiency and effectiveness of audits.

Natural language processing

Natural language processing (NLP) models driven by artificial intelligence (AI) – like ChatGPT and Google Bard – improve efficiency, ensure consistency, maintain quality assurance, and facilitate continuous assurance. These tools are particularly valuable when dealing with large volumes of unstructured textual data that require analysis and interpretation.

“On a global scale, AI is being integrated into data and analytics teams,” Moodley says. “Some assume that AI is solely for assurance purposes, but restricting ChatGPT’s use to data and analytics is akin to buying a laptop solely for the purpose of maintaining a grocery list. AI can be seamlessly integrated into various audit phases, including planning, control testing, and reporting.”

The role of digital assurance

Digital assurance encompasses technology, data analytics, and advanced tools to optimise assurance processes and enhance audit quality. It employs data analytics, AI, automation, and continuous monitoring to assess the accuracy, completeness, and reliability of financial and non-financial information.

“With these digital capabilities, auditors can analyse vast datasets, identify risks and anomalies, perform real-time or frequent audits, ensure compliance with regulations, evaluate cybersecurity measures, and provide stakeholders with timely, data-driven insights,” Moodley says. “This strengthens their ability to manage risks and make informed decisions.”

ChatGPT poised to revolutionise auditing practices

The potential of AI, particularly the development of private language models, is driving much of the transformation in the audit profession.

“BDO is developing a private NLP system to ensure data control while capitalising on the outputs,” Moodley adds. “Potential use cases include generating comprehensive insights into the content of working papers, with cross-references to the ISA (International Standards on Auditing). ChatGPT also offers technical accounting opinions with links to International Financial Reporting Standards (IFRS) documents or International Accounting Standards Board (IASB) standards.

“It can assist in risk assessment and management by providing access to risk libraries and aiding in the preparation of audit reports. The process involves curating publicly available data, building trust in its accuracy, and using it to enhance audit capabilities.”

AI-powered predictive analytics holds promise in enhancing data-driven modelling in business areas, such as credit assessment, enabling more accurate predictions by factoring in external variables, Moodley says. “AI can also assist in modelling risks comprehensively, improving risk assessment in areas like operational risk or provisioning for bad debts. Combined assurance, which typically relies on sampling, could benefit from AI-assisted risk modelling, leading to more accurate predictions and justifications.”

The move towards platform-based, data-led assurance is vital to address the widening gap between evolving business practices and traditional audit methodologies, Moodley explained. “The aim is to provide real-time, continuous assurance, and to transition from annual audits to more frequent assessments.”

Speeding up assurance

Traditional audit processes often involve sample-based methodologies, and findings are sometimes delivered up to six months after audit completion. In a world where data is doubling year-on-year, Moodley noted, this delay in providing assurance raises concerns about the relevance and coverage of audit findings.

“This ‘audit divide’ between evolving business practices and conventional methodologies requires a change in how assurance is approached,” he says. “The future of assurance is twofold: to become an integral part of the digital command assurance landscape, and to provide maximum coverage. Achieving this involves real-time and continuous assurance, possibly transitioning from annual audits to quarterly or even monthly assessments.”

From a business perspective, Moodley says, transformation is imperative due to the presence in South Africa of some of the world’s most innovative companies. “They are actively integrating cutting-edge technologies into their operations, leading to a radically different risk landscape. It’s a shift that directly impacts the field of assurance”.

Addressing the challenge of concentrating all risks into the hands of third-party entities is key. “One example of what can happen is China and its extensive use of AI to manipulate databases and large language models, and even predict criminal behaviour based on facial features. This is concerning,” Moodley adds.

He notes that the environmental and social implications of AI, particularly carbon footprint, requires organisations to assess their environmental impact and develop strategies for mitigation.

Bias and fairness

Biased data can lead to discriminatory or incorrect outcomes, Moodley cautions. To mitigate bias, it’s essential to scrutinise data sources.

He offers three examples of the potential pitfalls of biased data:

* In the US healthcare sector, biased data led to racial disparities in medical care assessment by AI systems, highlighting the risk of perpetuating historical biases.

* In criminal profiling, AI’s reliance on biased data resulted in disproportionate profiling of African Americans, even when white individuals had committed similar crimes.

* In workforce hiring, AI systems, influenced by historical data, favoured male candidates over females, undermining diversity in the workplace.

The people factor

When organisations resolve to implement an AI strategy, a critical challenge lies in reskilling their workforce.

“This can be a challenge, given the high demand for AI talent,” Moodley notes. “We must be cognisant of the importance of people and process in the profession’s digital transformation journey, and the need for effective change management, particularly when it comes to the workforce and established processes,” he says.

Panel discussion

SAICA project director for assurance, Angel Sithole led a panel discussion with Sam Bonser, product owner of verified analytics and Jason O’Shaughnessy, sales and partnerships vice-president, both at UK-based audit confirmation provider Circit; BDO director Michelle de Oliveira, and Nevellan Moodley.

The panel discussed how digital technology is streamlining tasks, improving accuracy, and enhancing data analysis capabilities.

Automating the creation of audit feedback reports and audit requirements is saving time, reducing the risk of errors, and consistently generating high-quality reports.

Algorithms developed to automate journal testing, especially for fraud and management override detection provide a more comprehensive analysis of journal entries, contributing to audit quality.

“Digital tools like Circit’s ‘verify transactions module’ pulls in data from client bank accounts for detailed cash testing,” Bonser explained. “Automating this process allows all sales invoice populations to match with payments received in the bank, improving efficiency and accuracy in testing. Bank transactions can be matched with accounts receivable, potential bad debts or valuation issues can be identified, and asset-related transactions can be comprehensively analysed.”

De Oliveira remarked that transformation driven by digital tools and technology is significantly improving teamwork collaboration among audit teams, and clients.

“Continuous learning is now a fundamental aspect of the profession, with auditors acquiring skills in data analytics and technology utilisation,” she added. “From looking back at historical data, auditors can now perform forward-looking analyses through predictive modelling and AI, enabling them to anticipate future risks.”

Despite the benefits, she warned, there are risks related to privacy, technology reliance, and complacency, which require vigilant mitigation strategies. “In the future, auditors will continue to adapt, focusing on critical thinking, independence, and engagement with clients to provide high-quality audits in an increasingly digital landscape.”

O’Shaughnessy discussed the shift from sampling to testing a larger volume of transactions using data analytics and real bank data, highlighting how this has been a game-changer for audit firms. He also touched on the evolution of technology over the past 20 years, including the impact of real-time bank feeds and disruptive technologies like mobile payments and accounting packages.

“The concept of open banking benefits auditors by providing access to trusted, real-time financial data directly from banks, improving fraud testing, completeness testing, and evidence collection. Some countries mandate open banking, while others are voluntarily adopting it, creating opportunities for auditors to utilise this data source.”

Moodley discussed the challenges and opportunities posed by non-authoritative materials from standard-setting bodies, particularly regarding the integration of digital tools in audit practices. “While these materials provide some guidance, they may not fully address the complexities of emerging technologies like AI. We need continuous engagement with regulators and standard setters to adapt to this new way of working.”

The panellists encouraged members of the profession to embrace innovation and experiment with new technologies, emphasising the importance of adapting to the evolving landscape. Digital transformation has promoted collaboration, continuous learning, extended stakeholder engagement, and improved reporting. These changes ultimately enhance the value auditors can provide to their clients and stakeholders.