When OpenAI launched ChatGPT, the AI-powered chatbot quickly gained attention from broad corners of the Internet. Everyone is wondering about the potential upsides or downsides of leveraging the tool in their industry.
Much of the buzz about ChatGPT is speculation because the technology is still new. However, it's not too early to consider the potential use cases and downsides of using ChatGPT for financial services.
Ways ChatGPT Could Be Used in the Financial Industry
ChatGPT was trained on language using both human and machine learning models. In supervised learning modules, the chat technology essentially listened to two humans talking and acted as a conversation assistant, learning by observation. In subsequent training, humans graded ChatGPT on its performance, effectively coaching it to perform better.
ChatGPT continues to learn from user interactions because users can upvote or downvote its performance.
A few potential use cases stand out when it comes to ChatGPT for the financial industry.
Customer Service
While complex questions still require a human touch, the industry is moving toward a day when AI bots can handle more customer service tasks. By using a bot to resolve issues quickly, customer service personnel can be reserved for those issues that truly require one-to-one service.
Automation
Much as you might ask Alexa to order a product for you, you could request a financial report be run via ChatGPT.
Easier financial reporting due to AI automation pays off in several ways:
- Dealmaking lifecycle: From due diligence through to closure, AI can revolutionize corporate dealmaking.
- Investor communications: ChatGPT could enable companies to send necessary communications to investors in an efficient manner.
- Progress reports and prompts: With multiple balls in the air, AI can provide helpful reminders and alerts to keep deals on track.
Informed Decision Making
As this technology advances, AI-powered bots may be able take over more complicated questions. Imagine a future where financial advisors and CFOs can tap into the knowledge of an AI to make decisions in real-time.
This may seem like a reach, but consider that AI can ingest significant amounts of data far faster than humans. When it comes to information analysis, pattern detection and forecasting, machines have the edge.
There are some weaknesses that currently prevent financial industry professionals from using artificial intelligence more broadly. However, the success of ChatGPT makes it clear: Soon, finance companies will be using the speed, power and algorithmic abilities of bots such as ChatGPT to react to market forces and make financial decisions.
Marketing
Marketers in the financial industry are always looking for ways to save time, allowing them to focus on higher-level strategy and messaging. ChatGPT's ability to analyze large amounts of data can provide valuable insights into customer behavior and preferences, helping to inform marketing decisions and improve campaign effectiveness.
ChatGPT can could be used to organize and structure information, introduce new concepts and ideas, and even generate a first draft of sales or marketing materials.
Weaknesses & Risks of ChatGPT for Finance
The full potential of these use cases is some time off, for now. OpenAI's CEO Sam Altman said as much in a tweet that remarked, "ChatGPT is incredibly limited, but good enough at some things to create a misleading impression of greatness. It's a mistake to be relying on it for anything important right now. It’s a preview of progress; we have lots of work to do on robustness and truthfulness."
AIs are not perfect, as Altman admits. Errors are perhaps the most obvious risk that comes to mind.
ChatGPT excels at conversation information gathering, such as answering questions and making suggestions. However, its answers aren't always truthful. The bot may suggest false information or merely something that cannot be verified easily.
Bias is an example of the unintended consequences that come with relying on a technology like ChatGPT in its present state.
As Bloomberg recounts, there have been notable incidents of ChatGPT writing sexist and racist content. The biased content was solicited by users to prove a point — that the AI learns and makes assumptions in ways that advantage some to the detriment of others — and the industry is far from a meaningful fix for bias in AI.
At present, Altman suggests downvoting biased content as a way to teach the machine inclusive values. Training machines on diversity reduces bias. To fully catch errors and minimize bias, companies must use safeguards alongside this technology.
It would be remiss to speak about warnings without mentioning cyber security.
The financial industry holds some of the most significant personal data, from account numbers to user addresses, Social Security numbers and other critical information. Before a tool such as ChatGPT could be rolled out broadly, there would need to be plans in place to protect consumer information. Not only is this ethically the right thing to do, but regulations also require it.
Ultimately, ChatGPT is the best iteration of a truly powerful AI chatbot capable of meaningfully contributing to the flow of information and streamlining operations across industries. Currently, AI tools assist the financial industry with discrete tasks including contract review software and data security. While these tools do a thorough job, it will take a long time to truly revolutionize the financial industry using ChatGPT and similar tools.
Nonetheless, companies should be watching these developments closely with an eye to the future. ChatGPT took many in the industry by surprise, and it's fair to say that an AI-powered future may be here before we know it. Certainly, companies should leverage AI tools to the full extent when doing so is safe, efficient and effective.
Discover why companies prefer DFIN when it comes to financial reporting software.
