AI’s Revolutionary Role

By Heath Morgan 

5 Things You Need to Know About ChatGPT, Generative AI, and Robotic Process Automation

The AI Revolution is Here

The generative artificial intelligence (GAI) revolution is here. Artificial intelligence (AI) technology has permeated various facets of our lives for years but when OpenAI launched ChatGPT in November 2022, the ease of adoption made use catch fire, and now GAI is everywhere, representing a true paradigm shift in technology in the industry. ChatGPT’s launch brought two concepts into the marketplace that have caused this rapid adoption rate:  1) they gave individuals the ability to “code” and generate prompts through natural language, and 2) they left its uses up to the imagination of individuals. These two concepts have effectively made ChatGPT an App Store for Generative AI and has opened the door to let users determine how they want to use the technology, and its list of uses grows each day. These use cases have expanded into development and adoption of two other popular AI software technology uses in the industry: robotic process automation (RPA) and conversational chatbots.

GPT & LLMs

ChatGPT is a Generative Pre-trained Transformer (GPT) and Large Language Model (LLM) that leverages the power of machine learning to produce various forms of content. Whether it’s text, images, audio, or synthetic data, GPT technology can generate it. Since ChatGPT’s launch, other established companies and upstarts have followed and offered their own GPT models including Google Bard, Chatsonic, and Jasperchat.

These GPTs, which operate in natural language format, can revolutionize company roles and responsibilities by creating content and first drafts for employees to then curate as needed. Its applications in the receivables management industry span from drafting email responses, creating compliance management system content including policies, procedures, work instructions, and training materials, to creating intricate legal pleadings and letters. It can also generate research on a new topic, allowing users to quickly get up to speed with subjects outside their usual expertise.

While the primary use cases are centered around public and free generative AI, companies are already exploring the utilization of enterprise GPTs and LLMs that are focused on controlled and proprietary data sets.

Robotic Process Automation

Another aspect of AI software technology that has sparked an incredible rate of adoption of robotic process automation (“RPA”). RPA technology enables software and bots to mimic human actions and automate repetitive tasks, freeing up time for employees to focus on more strategic tasks. RPA helps companies streamline operations, enhance productivity, and ultimately deliver better value to clients and consumers. Like GPT, RPA has existed in the industry for the past five years; recent players in the industry have brought the price point down considerably, to where this technology is affordable for small and midsize companies with a return on investment of upwards of 25 times. RPA has become especially popular in the legal process, where many repetitive human tasks are required to generate, print, and file lawsuits.

Conversational AI

The AI explosion has increased the receivable management industry’s interest in conversational AI, best known as chatbots and virtual assistants. This technology is reshaping consumer interactions and allowing companies to reimagine their call centers. These tools can understand, process, and respond to human language in a natural and conversational manner, improving customer service by providing instant responses and 24/7 support. As the Consumer Financial Protection Bureau’s (CFPB) June 6, 2023, issue spotlight on   revealed, all 10 of the top 10 banks are already utilizing chatbots for consumer interactions. While the receivables management association typically is slower and more cautious with adoption than banks and original creditors, debt buyers and collection agencies are rapidly looking to implement this technology to address one of the main pain points for any company – staffing.

All these software tools, including ChatGPT, RPA, and conversational AI, offer exciting prospects for the industry, especially in a difficult season for hiring and staffing. These technologies also present challenges that require a thoughtful and strategic implementation process. It is important for companies to learn about AI and how it can be implemented within their organization to stay relevant and thrive in this technological revolution.

Learn to Embrace Generative AI Technology

Embracing generative AI technology isn’t just about adoption; it’s about understanding its various forms and learning how to leverage them effectively within different contexts. Therefore, you can embrace generative AI even if it is determined that it is not a fit for your company right now. RMAI members should designate select employees or a committee to identify the right fit for the right type of generative AI.

Not all generative AI is created equal or should be treated equal. There are open-source and public models, like ChatGPT, where a company has no control of the data once it has been submitted. There are enterprise and proprietary AI that give customers confidentiality and the right to delete data. There are rules-based systems, which program out every generative output, and there are sophisticated black-box systems, which can be more creative in their output. There is manned generative AI, which requires a human to prompt and program for the output, and there is unmanned generative AI, which allows for AI to self-code outputs and run autonomously. This means that generative AI isn’t a monolithic entity — it’s a spectrum with varying degrees of capabilities, uses, complexities, and required safeguards.

Understanding these nuances will enable businesses to create strategies to utilize AI to its full potential, taking into account their unique operational needs, objectives, and safeguards, and vital for anyone wishing to leverage the potential of generative AI technology.

Companies in the industry are already using ChatGPT and other generative AI for policies, procedures, work instructions, training materials, quality assurance controls, reports, marketing materials, email responses, legal pleadings and responses, new technology adoption, business case studies, and many other company roles that involve content creation, or require research.

The other important aspect to understand about this technology is that it is here to stay, and will be adopted by both industry vendors, as well as consumers very soon.  Even if RMAI members decides not to utilize ChatGPT or other public LLMs because of the risk to consumer, confidential, and proprietary data, there is still value to learning how prompts work and potential use cases for the technology to be ready when there are LLM products that meet your standards for confidentiality.  Additionally, there are already vendors in the industry utilizing GAI for their services, as well as several companies marketing consumer based generative AI chatbots to interact with companies and outsource their customer service experience.

Implement AI Safeguards in Your Company Today

Because of this adoption rate by vendors and consumers, it is imperative for RMAI members to look to implement AI safeguards within your company. This includes creating an AI policy, creating an AI committee, and creating an AI risk assessment process for implementing new AI technology.

Creating an AI policy—establishing a framework to manage the use of AI and protect consumer data, and confidential and proprietary data—is a must. Companies should treat public GPT like ChatGPT with the same caution as they would treat social media platforms, given that the current versions provide no confidentiality or protection over any data you share. To be clear, companies should not provide any confidential, proprietary, or confidential data to public or private LLMs that do not offer confidentiality to your company standards, and that should be the fundamental safeguard when utilizing this technology.

Company polices should clearly define who within the organization is allowed to use AI tools including ChatGPT. This helps maintain control over how these tools are used and ensures they are used responsibly. Along with this, companies should prepare for all prompts, queries, and output to be requested by clients and regulators, and should create a transparency log for such inputs and outputs.

Establishing comprehensive AI safeguards can ensure that businesses use AI responsibly while maximizing its benefits, and ensure that companies are investing in the right type of generative AI tools.

Adopt a Future Forward Look at AI

As you adopt your AI policy, two additional areas to consider are one, how will AI be regulated in the future, and two, how will consumers and other actors use the technology?

As AI continues to advance at a rapid pace, so does the regulatory scrutiny surrounding it. Both the CFPB and Federal Trade Commission (FTC) have released statements about the use of AI in the financial services industries. It’s important for companies to familiarize themselves with these regulatory guidelines and be prepared for the changes these may entail, and use those to guide investment and development.

For instance, in the CFPB’s issue spotlight on chatbots, concerns were raised about the use of AI which resulted in risk of noncompliance with federal consumer protection laws,  diminished customer service and trust when chatbots reduce access to individual human support agents, and harming people including difficulty in resolving consumer disputes, providing inaccurate, unreliable, or insufficient information, failing to provide meaningful customer assistance, and being vulnerable to impersonation and phishing scams. The CFPB’s issue spotlight on chatbots is a must-read for anyone implementing conversational AI bots and RPA. Most importantly, it lays out a safe path for companies to get started.

Another jurisdiction to look to is the European Union (EU), which recently passed an EU AI Act Resolution. This legislation categorized high-risk AI systems, including those that could harm people’s health, safety, fundamental rights, or the environment. The EU AI Act also calls for companies to be transparent in how their AI systems work, requires risk assessments before deploying them, gives consumers the rights to challenge decisions made by AI systems, and bans real-time biometric identification and social scoring systems. It would not be a stretch to imagine similar legislation in the United States in the future, as both the Democratic and Republican parties have expressed interest in the regulation of AI.

Another concern for the RMAI members is consumers and bad actors using generative AI against company systems and employees, both through online chatbots and conversational AI. Expect to see call baiting bots, dispute bots, and other virus agents engaging with your website and agents in the near future. Additionally, expect to see an uptick in scammers using generative AI to increase the attacks against your system and people through phishing, fraud, or ransomware.

Companies must proactively incorporate potential regulatory concerns into their AI policies and should allocate adequate resources to prepare for potential harmful uses of generative AI against their company.

ChatGPT is Just the Beginning, and That Should Be Exciting

We are only at the dawn of the generative AI revolution. ChatGPT, with all its impressive capabilities, is only the first version of what generative AI and LLMs can potentially offer. Current concerns around safeguards and privacy are indeed important, but these concerns should not deter companies from exploring and developing within this field. Future versions of generative AI and LLMs will have enhanced proprietary, confidentiality, and data deletion rights.

We are only a few months away from enterprise LLMs where companies will be able to purchase data for their LLMs, as well as being able to upload their own data into their own LLM with privacy and confidentiality. The uses of these within the industry will be endless, from your own internal CMS bot to train your employees to litigation bots to write briefs for specific judges or against specific FDCPA attorneys. Hundreds of GPT-based products will soon flood the market, each with its unique benefits and drawbacks.

Just as GPT technology continues to increase its adoption rate, RPA and conversational AI continue to gain traction in the industry as new use cases are being implemented every day, the price drops, and ROI increases.

The advent of ChatGPT, generative AI, robotic process automation, and conversational AI herald a new era in the industry landscape. An exciting yet challenging future awaits us, especially as our economy shifts from a disruptor collaboration economy.

The good news is that the receivables management industry has long been built on collaboration as a necessity to survive, and collaborative companies look for ways to work together with strategic partnerships to streamline innovation and inefficiencies. It’s an exciting journey, and we are only just beginning.