What is an algorithm? Let’s demystify Algorithms and Artificial Intelligence (AI) has been saved
Perspectives
What is an algorithm? Let’s demystify Algorithms and Artificial Intelligence (AI)
Benefits and challenges for auditors to consider
Algorithms may benefit individuals in many ways, from helping us determine what products to buy to providing us the fastest route to work—but algorithms may also improve efficiencies on a larger scale with finance automation. The use of algorithms poses unique benefits and challenges for business leaders in finance, but before we dive into them, let’s cover the basics.
Understanding the types of financial algorithms
At its core, an algorithm is a process or set of rules to be followed in calculations or other problem-solving operations. Common examples of rule-based algorithms include if-then statements, which can often be found in simple spreadsheets. For auditors, these three real-world examples highlight some of the use cases along the spectrum of algorithms, including automation and artificial intelligence, in today’s finance operations:
Future of Work: Ways of working in uncertain times
${column1-large-text}
Robotic Process Automation (RPA)
Handles high-volume and repetitive tasks using rules-based "bots.” These bots can extract key data from various documents, such as bank statements, checks, and invoices, and summarize cash balance reconciling items in a single report.
${column2-large-text}
Algorithmic Trading
Can drive which position to enter and close using a predetermined set of rules. Traders design these algorithms to generate profits or rebalance portfolios to manage risk.
${column3-large-text}
Artificial Intelligence and Machine Learning
Interprets data, recognizes patterns, and generates outputs - in certain cases mimicking human intelligence. Generally, the intelligence in artificial intelligence stems from the underlying algorithm's ability to dynamically adapt to training data and improve the model.
${column4-large-text}
${column4-title}
${column4-text}
How do these algorithms differ?
Beyond understanding what algorithms are, auditors may also need to understand the characteristics of an algorithm, specifically in terms of intelligence, complexity, and type. These characteristics vary along a sliding scale, influencing the level of inherent risk associated with a given algorithm. Become better prepared to identify and assess algorithms implemented in your business process by gaining knowledge around these algorithm characteristics:
Finding algorithms in finance systems
Identifying algorithms might seem simple in theory, but they are often hidden in larger business processes, applications, and software. In finance and accounting, these algorithms are embedded within business operations to create outputs for documentation and financial reporting. While finance and accounting personnel depend on these outputs, they may not be aware of the impact from algorithms used in upstream business processes.
To identify algorithms, auditors may consider a consistent line of inquiry when conducting walkthrough discussions with finance, operations, and operational support personnel. These questions may be tailored or expanded upon based on the unique nature of the business or process under review. Once the audit team identifies the algorithms used across the organization, they can more accurately develop the risk assessment, planned controls, and substantive procedures to address identified risks of misstatement.
Risks associated with algorithms
While algorithms can increase operational efficiency, they also pose potential risk that may impact account balances, estimates, or disclosures within financial statements. To assess an algorithm’s risk level, auditors can analyze the quality of the input data, possible assumptions or bias linked to specific inputs, methodology behind the algorithm, and the IT environment in which it resides. By assessing the risks associated with algorithms, the audit team can work with subject matter resources to design an appropriate audit response.
Contact Us
Get in touch
The services described herein are illustrative in nature and are intended to demonstrate our experience and capabilities in these areas; however, due to independence restrictions that may apply to audit clients (including affiliates) of Deloitte & Touche LLP, we may be unable to provide certain services based on individual facts and circumstances.
Recommendations
NYC local law 144-21 tackles AI bias in employment decisions
Preparing for new rules under New York’s legislation
Reduce AI risk and promote AI trust
The remarkable potential of artificial intelligence (AI) may be tempered by growing challenges to data privacy, risk mitigation, and cybersecurity—in short, to trust and transparency.
