The launch of OpenAI with ChatGPT in late 2022 ushered in the age of generative AI for the masses and with it, a whole new world of innovation, opportunity … and risk. Companies worldwide have had to quickly assess any threats to their operations that these new tools and balance innovation against responsible use.
The productivity gains enabled through generative AI had to be balanced against concerns of security, and ethnical use. Corporations like Experian that specialize in big data and information technology were quick to not only institute internal policies for the responsible use of generative AI, but also establish many of the best practices that its clients should consider in using the tools.
Looking back at the past year of generative AI’s global impact, Digital Journal asked Shri Santhanam, executive vice president and general manager of global analytics and AI for Experian, to share his insights on what every company should be considering regarding its use.
DJ: What is the most promising takeaway you are seeing companies embracing when it comes to the use of generative AI?
Shri Santhanam: Without a doubt, Experian and our clients are seeing how much generative AI can drive innovation within an organization. For example, we are seeing that software engineers and development teams gain much more efficiency and productivity in the coding process. Generative AI tools can accelerate product development, testing, and quality assurance which, in turn, helps software engineers broaden their skill set.
This capability dramatically fast-tracks the exploration of new ideas, business models, experimentation, and internal decision-making. We think this type of “grassroots innovation” must be embraced, nurtured, and put into operation at some level to remain competitive and activate the full potential of your team.
DJ: After a year, are there any early warning signs or advice you would give businesses working with generative AI?
Santhanam: This is a good question because like any new technology, there is the danger of spending too much time in research and not enough time developing products and services that drive productivity or profitability. A real danger is going down the proverbial development rabbit hole, exerting time and resources on efforts that will not pay off.
To avoid these traps, businesses should start by identifying the specific problems that their customers or internal operations are experiencing, and then work backward to ascertain how technology can be applied to solve those problems. When you know your end goal, you can make targeted investments in time and resources to develop solutions!
A good example of this is improving automated customer service. Generative AI and machine-learning technology can help support answering questions from consumers and getting them answers them on business-related topics or processes. At Experian, we speak with hundreds of consumers every day through both live conversations and our websites, helping them with credit education. So, we ask ourselves “how can generative AI improve this dialog, make it faster and help consumers get the best financial and credit education?” This question drives our innovation.
DJ: Much has been written about the dangers of generative AI in fraud and ensuring data privacy and protection. How is Experian addressing this?
Santhanam: There are a few ways to look at this very issue. First, it’s important that we follow and adhere to regulations. AI is not new to us. For the past two decades, we have been harnessing the tools of AI to be more specific and surgical in lending in different applications and models. These models run on data, and Experian has massive amounts, so a big part of our focus is regulation and safety.
Another way to look at this is ensuring your own workforce understands that data protection starts with them. All companies that use or hold data on their customers need to recognize that this data must remain secure to ensure trust in the brand. This safeguarding starts with a company’s workforce, who must be trained on the proper use of generative AI tools and – until they are – corporate access to those tools should be limited.
Finally, we must be conscious that if generative AI can enable corporate innovation, it can also give fraudsters a powerful tool to carry out more sophisticated attacks more rapidly. This reality makes the need to protect corporate and consumer data an absolute requirement. Companies should strategically and seamlessly assemble a variety of multilayered detection capabilities that include authenticated identity data, device-risking, email and mobile risking, behavioral biometrics, document verification, and other risk signals to instantly spot early indicators of potential fraud.
DJ: With these technical advancements moving so quickly, it can easily seem daunting to keep up. Any advice?
Santhanam: There’s no question generative AI continues to drive innovation of new products and services at hyper-speed. It’s important that companies maintain a flexible approach to AI-related investments. An inherent risk is that long-term investments in developing offerings that employ AI may lead to future obsolescence as the technology continues to advance. A prudent approach is to make short-term investments that easily migrate into long-term ones, such as investing in partnerships and productivity software that can be easily and cost-effectively deployed as well as focused tuning of open-source models for specific use cases versus a large investment in foundational models.
But make no mistake, generative AI is a disruptive technology that will forever affect how companies manage their data and develop their products and services. It is important that the right amount of attention is given to this technology to ensure the responsible use of the tools, mitigate the risks.
