3 minute read
Generative AI
Understanding the rewards (and risks) for businesses
BY AVI PATEL
Prompt: Write a 500-word article about how businesses, especially banks, are using artificial intelligence (AI).
Response: This article wasn’t written by AI, but it probably could have been.
NO DOUBT YOU’VE HEARD ABOUT the excitement around generative AI, which can be given a prompt and develop new content, which could be text or visual presentations, from existing information.
There’s a lot to be excited about. Who wouldn’t want to take advantage of technology that can summarize a 20-page document into a few paragraphs? Think of the time savings!
Generative AI using large language models can be applied to a wide range of initiatives:
Summarization, such as taking the transcript of a recorded meeting and developing a concise synopsis
Classification, including conducting sentiment analysis of thousands of individual survey comments
Content generation, such as basic marketing content and articles
Chatbot functions, which leverage a question-and-answer format to draw information from a company’s knowledge base
Translation to different languages
Data analysis to answer specific queries based on large data files
The possibilities can seem endless, and they need to be balanced against what’s appropriate for a business, especially in a highly regulated industry like financial services. For instance, customer data should never be input into a generative AI platform, as that information should remain private and under the control of the company.
Other uses of generative AI are fair game. For instance, generative AI can be used for research projects that involve reviewing publicly available reports. AI can expedite and accelerate this process, distilling large volumes of information from multiple sources into a more concise summary that includes the key information. For a community bank, generative AI can be used to read through customer commentary and distill it down to a few key messages for bank leaders to review and act on.
Like all technology, there are some drawbacks. With generative AI, one area to be cautious of is “hallucinations,” which is when the tool provides information that may be false or misleading. Some AI tools are designed to answer your prompt or question, no matter what, even if it doesn’t have sufficient data for a factually sound response.
As you explore using generative AI, be sure to consider all aspects of how it could impact your business, including not just technology but also legal, compliance, risk management, and how might be perceived by your employees.
also critical to understand the data protection and security aspects of any AI tool, as well as whether any files or data you provide be incorporated and used to help “train” the AI platform. The use of generative AI to enhance internal business operations and customer experience is just beginning, and as with all new technologies it’s important to let the overall business strategy drive adoption and implementation of artificial intelligence.
Avi Patel is chief data scientist and chief marketing officer at Fulton Bank.