Ethics, accountancy, and AI-powered tools

Artificial intelligence is proliferating at an unprecedented rate. And as AI and its use cases evolve, so does the way finance professionals practice. The rapid rate of change also creates new ethical challenges within the profession.

With increased capabilities such as predictive forecasting or AI smart assistants, there is a greater opportunity for innovation and growth, which provides greater capacity to add value. But it is important to not let this perspective change or diminish your standards.

Exploring both the potential benefits and risks of AI-powered tools and what we can do to balance the two is vital from an ethics perspective.

AI-powered tools for efficiency, risk management, and forecasting

You are likely already discovering numerous ways AI-powered tools benefit your professional practice. For example, automating routine tasks and admin, such as invoice categorisation and reconciliation, which can be handled by AI tools. Offloading admin-heavy tasks to machine learning and AI allows you to focus on better ways to advise clients and employers and improving performance.

Appropriately conducting risk assessments is key to the success of any business. AI has vastly improved the speed and capability of conducting risk assessments. Software closely examines anomalies in expense claims or procurement data and reviews an exhaustively large volume of data that would take a person infinitely longer to review. AI tools can detect errors and even fraud, thereby improving internal controls and supporting stronger governance.

Financial forecasting can be of immeasurable value to any business, particularly in times of market volatility and global instability. AI software can process large volumes of data to deliver financial forecasts at record speeds. Of course, it has to be cautioned that the quality of any output is inextricably linked to the quality of the input data. Always verify the findings.

By integrating AI tools into a client’s or employer’s finance management software, it is possible to obtain comprehensive predictive insights, saving time on the heavy lifting of manual data analysis and allowing more time to consider the implications of the results.

Ethical risks of AI-powered tools

The benefits of AI-powered tools are plentiful. However, despite the increases in efficiency, accuracy, and availability, it is still necessary to take a closer look at the risks.

The CIMA Code of Ethics requires that CIMA members “maintain professional knowledge and skill at the level required to ensure that a client or employing organisation receives competent professional service, based on current technical and professional standards” (Code of Ethics R113.1(a)). So, while it is valuable to keep up to date with the latest developments in AI to provide the best service possible to clients and employers, it is also wise to be wary of competence atrophy. An overreliance on AI software can risk a decline in professional competence.

Generative AI often requires access to vast amounts of data to improve output quality. It is important to remember that AI doesn’t “learn” as humans do but, rather, is trained to produce a specific output using an algorithm that predicts the type of output sought by the user's prompts. Ideally, the more data the software has, the more accurate its output is relative to the prompt's precision. But in reality, the quality of the data input is as important as the volume. Using AI for data analysis can also create data security risks, as many AI software systems use cloud-based AI services that are often targeted by malicious actors such as hackers.

Be mindful of the nature of the data being fed into AI software at any time. Do you know where and how it is stored? If not, you may wish to find out.

Ethical integrity in AI integration

Striking a balance to maintain an ethical approach to implementing AI-powered tools is vital. The ethical considerations are not just about the right and wrong of using AI, but also about understanding what you need to know to ensure that professional standards of practice are not compromised by a desire to utilise AI without limits.

When implementing AI-powered tools, you must understand the software's limits, how it works, where data is stored, and, where possible, put safeguards in place to protect sensitive data.

Remember to verify outputs and to maintain your professional competence. Key questions to ask yourself when receiving a recommendation or forecast include:

• Have I analysed and assessed the results to ensure the recommendations are accurate?

• Are there appropriate levels of oversight and transparency regarding the use of AI software?

You spent years developing the knowledge and skills necessary to practice to the highest professional standards as an accounting and finance professional. This responsibility cannot and should not be delegated to AI.

AI should support, not supplant.

Explore more insights on AI and ethics, and read the report Making good decisions: Building trust in a digital world.