Chat GPT – What Generative AI Means for Finance Teams

Chat GPT, you can’t escape it. Whenever something billed as transformative comes onto the market, there’s always a flurry of excitement around it – whether it’s the first iPhone launch, Google Maps, or if you’re in Reading, the opening of a new Wendy’s restaurant… But it doesn’t matter how much marketing you put behind a product, if it doesn’t live up to expectations, then the flurry quickly turns to tumbleweed.

When Open AI launched an early demo of ChatGPT at the end of 2022, social media lit up with tales about remarkable it was. Why? Because it worked. And because it was simple to use. Within five days it had over a million users. To begin with the use case seemed limited to article generation, and it probably still is best known for its human-like written “voice,” but it’s easy to see how that can be used in finance scenarios. Generative AI uses machine learning algorithms based on existing data to create new data, so finance teams could leverage that information for contracts and investor or supplier communications for example. For finance teams still mired in manual processes, with an abundance of paper in the room, ChatGPT may seem like an irrelevance. But that’s an assumption that’s probably misguided, and here’s why:

Forecasting

The bedrock of company stability starts with knowing current status and how that relates to future prospects. The speed at which AI can generate answers and combine different streams of data, will accelerate as Generative AI collaborates with traditional AI to create reports, analysis and crucially, provide recommendations.

Risk and Compliance

Large language models like ChatGPT are very good at identifying patterns, and by using GANs, for example, can spot anomalies and potentially suspicious transactions. They can also act as an early warning system for compliance breaches, meaning that organisations can act fast before it becomes an audit issue.

Core Process Transformation
As the technology develops, we'll start to see a revolution in some of our core processes too, as Generative AI combines with standard automation to go beyond efficiency and transform areas like invoice processing, even if those invoices are complex and unstructured. Currently many AP teams already use AP automation for for speed, accuracy, workflow and data extraction, and with the addition of Generative AI, can read and interpret contracts and invoices like a human would, but at speed.

But there are Challenges

The dynamism that’s driven Generative AI usage is also its biggest weakness. It’s very easy to experiment with, which raises security concerns and potential leaks of confidential data. Another problem is that Generative AI can produce very convincing wrong answers, which if acted on, could lead to some very expensive mistakes or embarrassments.

Try it Out

The trick to understanding it, is to try it. As the barriers for use are so low, anyone can do it. Start with something small and discover the practical applications fit for your business.

Generative AI is here to stay

We’re just at the start of the next revolution of AI and it’s exciting to think about where it will go from here. As new iterations become ever more powerful, finance leaders of today need to grasp an understanding and foster an atmosphere of curiosity and learning within the rest of the team. The power of the human brain is remarkable, and alongside Generative AI, the possibilities are endless. But remember, just because you CAN automate something, doesn’t mean that you should. So as with traditional processing and automation, human insight, human intuition, and human knowledge is here to stay too.