Generative AI (GenAI) has taken the world by storm. We’re not just referencing the usual tech ecosystems and clusters, but mainstream consumer use as well. ChatGPT and other large language models (LLMs) have permeated the everyday lives of people everywhere, not just tech employees. One industry where the technology has enormous potential is financial services and fintech. Rather than look at horizontal use cases across industries that can be applied to financial services – coding, marketing, etc. – we’re focusing on actual vertical use cases that have appeared within various fintech verticals today.

WealthTech

In WealthTech in particular, financial institutions struggle with data inefficiencies, which cause information and data to go to waste. Data is often siloed and inaccessible, or lacks governance altogether. But with the help of Generative AI, wealth managers will have the ability over the coming years to provide more accurate information to clients and identify trends or patterns in collected data. In addition, Generative AI has ignited an industry-wide desire to create more transparency in the financial markets and offer a digital-first experience for investors and institutions alike.
As financial services become more complex with new assets and frameworks, Generative AI can play a critical role in expanding financial education for potential customers while also bolstering corporate training programs and assisting employees on the job with answers to potential inquiries. Generative AI has also been central in enabling wealth managers to be able to incorporate the use of chatbots and robo-advisors into the client experience.

Use Cases:
Morgan Stanley’s advanced chatbot effectively collects and summarizes the bank’s depository of data that would otherwise be dispersed in the cloud.

LTX launched BondGPT which answers bond-related questions. Old AI tools struggled with pulling together all of their data to create one simple user interface as compared to Generative AI. Banks will follow suit in creating similar technologies: JPMorgan has already applied to trademark a product called IndexGPT which selects investments for customers using Generative AI.

InsurTech

Insurance underwriters have experienced notable challenges when it comes to navigating business risk research in a timely manner. Thankfully, the use of generative AI models can assist insurers in proficiently collecting and examining large amounts of data, replacing current systems in insurance that force insurers to navigate through piles of data. While insurers previously attempted to use machine learning and predictive AI, generative AI uniquely scales underwriter’s abilities to comprehend information at a quick pace, especially within complex underwriting processes such as commercial insurance that includes more data and context.
By harnessing the power of Generative AI, insurers will undertake a revolution in the industry by being able to offer more competitive and personalized pricing for customers.

Use Cases:

Viola portfolio company, Planck, has developed a cutting-edge technology built on Generative AI that allows underwriters to analyze huge amounts of data, build complex insights, and identify patterns that were previously inaccessible. Carriers can automate and streamline the underwriting process, enhancing consistency and accuracy for better customer service and optimal organizational success. Planck’s platform is an AI underwriting assistant that generates actionable insights across the insurance lifecycle to power the commercial insurance strategies of the future.

Payments

With attempted fraud skyrocketing 92% from 2021 to 2022, the payment industry is faced with an immediate need to adopt solutions that help protect customer data. Even though predictive AI did have an impact on fraud detection rates, Generative AI has proven to be a powerful detection system against identifying fraudulent activity. The large language models (LLM) behind Generative AI are able to analyze large pools of data, meaning it can be used to identify potential fraud patterns, pre-empt fraud tracks, and filter away low-likelihood fraud alerts.
Generative AI will also speed up the ability for payment providers to implement voice-activated payment systems which will add another layer of security in fraud detection capabilities by analyzing vast amounts of voice data to identify speech patterns and improve voice recognition accuracy.

Use Cases:

Payment providers such as Visa, Mastercard, PayPal, and Bank of America have used Generative AI to mitigate fraud. While it is still early and difficult to pinpoint exact results to date, the largest companies in the payment space have dedicated sizable investments into Generative AI with fraud detection capabilities.

Lending

Generative AI has begun to help lending companies improve the workflows for loan processors by streamlining the loan underwriting process.
Automation significantly decreases the time and expenses associated with loan processing and produces higher rates of accuracy. By reducing processing times and creating more accurate measures to tailor personal loan offerings, lenders will attract future borrowers and increase their customer base. Another use case of Generative AI in the lending vertical is unified data management and document standardization. This is especially important in a world with 10,000+ banks and credit unions, pay stubs can come from hundreds of thousands of payroll systems, all with different formats.

Use Cases:

AIO Logic is a commercial loan management platform using Generative AI for risk and rate determination and custom loan structuring.
Ocrolus uses Generative AI to offer digital lenders document automation. Ocrolus’s technology reduces both document analysis times and overload on human analysis as compared to traditional AI, which struggled to accomplish these tasks. As a result, lenders will be able to scale their operations and lending capacity by accelerating document analysis and cutting human workloads, which will drive company growth.

CFO Stack

In the current economic climate, CFOs are prioritizing cost management and sustainable balance sheet management to drive profitability. Generative AI will play a pivotal role in the modern CFO stack by automating tasks that traditionally demanded the finance team’s time and resources. The technology will be adopted by many businesses to transition CFO’s focus from solely crunching numbers to becoming strategists for company growth and expansion.

Use Cases:

Brex has launched advanced Generative AI-powered tools for CFOs. The tools help provide finance teams with relevant information regarding corporate spend and answers to critical business questions related to budgets and spending patterns in real time through a chat interface. With these recent advancements, CFOs will even be able to create custom graphs to help visualize spending patterns, which is something previous AI tools could not accomplish.
In addition to Brex, Auditoria implemented Generative AI to its technology stack for Smartbots trained in finance language. Auditoria’s Smartbots intelligently automate complex finance workflows while collaborating and engaging with the company’s customers, suppliers, vendors, and internal stakeholders through conversational email. These capabilities have been proven to increase finance teams’ speed and accuracy through their integration with many financial applications.

Banking

The banking industry is entering an era of hyper-personalization where customers seek out tailored financial solutions and demand services that are convenient, simple, and unique to their own financial goals. Especially as there are an increasing number of banking touchpoints every single year, customers expect top-quality product offerings from institutions. Generative AI will help banks analyze large volumes of customer data and leverage it to learn about customers’ preferences and financial behaviors, allowing banks to offer more personalized and effective product recommendations for end users.

Use Case:

Another Viola portfolio company, Personetics, incorporated OpenAI’s models into its existing AI processes to enhance its personalization solutions for financial institutions. As key to their financial data-driven personalized engagement platform, the combined AI models increase Personetics products’ time-to-value, accuracy and customer experience. For example, their ‘Enrich’ product helps customers make sense of their transaction data by cleaning, enriching and categorizing it whereas their ‘Engage’ product improves customers’ financial well-being while increasing engagement and sales. By leveraging GenAI, Personetics will make strides in reducing analysis time plus better analyze customer feedback of each hyper-personalized insight to increase the level of personalization and customer experience.

Risks and Mitigants – What Fintechs Need to Look Out for When Implementing GenAI

While the emergence of Generative AI brings tremendous opportunity to the financial services industry, it also introduces a number of risks that are critical to address. Namely, the limited context of data in LLMs, which may produce “hallucinations,” or answers that are not accurate 100% of the time. There are, however, a few ways to mitigate the risk of “hallucinations” and ensure data output is accurate. This includes using zero-party data and first-party data and preparing the data through pre-processing to ensure all information is accurate.The “black box risk” is another primary risk for Generative AI users, which refers to the lack of transparency and interpretability in Generative AI models. Companies can have difficulty understanding how decisions are being made. In financial services, this opacity can pose significant challenges in terms of regulatory compliance, accountability, and risk management.
One solution to addressing “black box risk” is using a chain of thought prompting processes to generate proper reasoning (such as a step-by-step process). Additionally, companies that adopt Generative AI tools should develop frameworks and rigorous model validation processes to ensure that there is accountability throughout the implementation process.

As the financial services industry continues to develop and adopt new practices, striking the right balance between innovation and risk mitigation will be key to unlocking the full potential of Generative AI in reshaping the future of finance.

The above list of risks is in no way exhaustive, and opportunities are also growing and shifting as this vertical is taking shape.

Stay tuned for Part II of Viola’s series on the intersection of GenAI and Fintech, and be in touch if you are working on something compelling in this field.