Artificial intelligence (AI) is transforming nearly every aspect of our world. Different types of AI are rapidly becoming integrated into the business and finance world, often with mixed results.
There are both advantages and risks to adopting (or not adopting) AI tools for financial work. Companies integrating AI more quickly are seeing strategic advantages, but they’re also dealing with increased risks. The increased risks underscore the need for vigilance regarding new technology, both internally and from external bad actors. We’re also seeing some studies indicating that AI is not decreasing the workload on human employees as much as anticipated.
Agentic AI, Generative AI, and Machine Learning
There are several different types of AI tools, and we need to define some terms before talking about how AI is impacting finance.
When most people today hear the term “AI,” they think of Generative AI (Gen AI). This type of AI generates text, images, or some other output in response to a prompt. Large language models (LLMs) like ChatGPT are examples of this type of AI.
Machine learning (ML) is a type of AI that examines data and looks for patterns to improve the way the software performs tasks. In finance departments, Business Process Automation (BPA) software uses ML to “learn” your company’s way of managing finances and become more useful the longer you use it.
Gen AI and ML are both key parts of Agentic AI. Agentic AI is designed to make decisions and take action. It does not always require input from a human user to function. Examples include virtual assistants, automated workflow management tools, and responsive customer service bots.
1) CFOs Gain A Strategic Advantage
In companies that have already extensively integrated AI into their business models, a global survey of executives showed that 73% “believe they are gaining a strategic advantage” by using AI. This report, Executive Perceptions of Artificial Intelligence (AI) Opportunities and Risks: A Global Analysis, included a survey of 1,735 executives representing eight major industries, non-profits, universities, and government agencies from around the world.
This survey also found that companies that are slow to implement AI are starting to fall behind. North American companies are also far less likely to prioritize AI transformation than Asian and South African companies. Though AI is seen as risky, not investing is also a risky decision that can put companies at a strategic disadvantage.
2) Weighing the AI Risk
The global survey referenced above found that even though executives at companies with high levels of AI adoption gain a strategic advantage, 63% also report that AI risk is a top focus for them. As AI becomes a bigger part of the business world, it also becomes more urgent to deal with the risks that AI poses.
AI poses several risks to companies. Two of the most concerning risks are related to company reputation and cybersecurity. If AI agents report incorrect data, offend customers when interacting with them, or leak private information, those errors can damage a company’s reputation. AI can also amplify cybersecurity risks from hackers using AI or from the company’s AI leaking protected information. Other potential risks include regulation uncertainty, AI violation of intellectual property, AI’s ability to enforce compliance, and AI hallucinations that generate inaccurate outputs. As companies integrate AI into their business models, they must anticipate and take steps to mitigate these risks.
3) Increased Need for Vigilance
The many risks of AI increase the need for vigilance and training to recognize AI-related issues. Two of the biggest risks come from outside threats leveraging AI and AI hallucinations. From the outside, AI can be used as a tool by malicious workers. Hackers and scammers can leverage AI for attacks on your business, including deepfakes where scammers impersonate clients or executives.
One of the major risks of AI hallucinations is that the AI is confidently wrong. Because the outpuast carries a note of confidence, human workers might accept the information at face value. For example, certified professional accountants might see clients trying to claim unallowable business deductions because an AI tool told them it was okay. Or an AI agent might make up data when generating a report, which can lead to inaccurate reporting, forecasting, etc. Employees must be trained to double-check AI output.
4) Human Workload Increases
Only 16% reported that their workload decreased. The majority (53%) of respondents said using AI has not impacted their workload at all.
With the expectation that AI will make work easier, many employees report that their managers’ productivity expectations have increased. This puts increasing pressure on employees. About 37% are already reporting AI fatigue, and 14% of workers say that AI has made them more dissatisfied with their jobs.
Another global study sheds light on why there are so many productivity issues with AI. The report “Beyond Productivity: Measuring the Real Value of AI” stated that “roughly 37% of the time saved through AI is being offset by rework.” AI tools help people work faster, but instead of using the time they save “to focus on higher-impact work,” employees have to spend that time “correcting low-quality AI output.” Training employees to use AI tools effectively, measuring productivity by outcomes, and redefining job roles in light of AI’s impact can help counter these issues.
5) Consider Starting with BPA
With all the potential downsides to using AI, as well as the relentless pace of new changes to AI tools, it can be tempting to delay integrating AI into your company or your individual work. However, there is also a risk in ignoring AI completely. Companies that don’t use AI are starting to find themselves at a strategic disadvantage. Also, the outside risk of malicious workers using AI against your company won’t go away simply because you don’t use AI yourself.
At the same time, don’t feel like you have to start using AI just because. Make sure that there’s a purpose behind AI integration and that your company is implementing AI thoughtfully, with proper training for employees.
One way to start using AI in your company is by implementing BPA software with integrated AI tools. NextProcess software integrates machine learning to automate financial processes. The more you use the software, the better it understands your company’s preferences and policies to ensure accurate auto-processing. Our software modules provide an end-to-end solution to power and simplify the entire procure-to-pay cycle. Contact us today to schedule a free demo.
