Solomon Philip is Shift Technology’s Head of Market Intelligence
ChatGPT is everywhere. There is no denying that generative artificial intelligence (AI) is the topic of conversation and that opinions about what it really means for business, education, and entertainment run the gamut. Naysayers insist it’s just a fad, or that it's something fun to play around with and create cute little stories and pictures to share with your friends to make them laugh.
Doomsday prophets see it as one more sign that the machines are truly becoming sentient and that the human race’s days are numbered. Ask 10 people on the street what ChatGPT is all about and the impact it will have on their life and you're likely to get 10 very different answers.
But the reality is that generative AI is here to stay, and like most emerging technologies, its benefit depends entirely on how it gets harnessed.
At Shift, we keep a careful eye on those technologies that have tremendous potential for transforming the insurance industry, so it goes without saying that ChatGPT has not only been on our radar, but also on our roadmap for quite some time. In fact, we’ve been incorporating Large Language Models (LLMs) in our insurance decisioning solutions since 2020, giving us a head start in understanding the benefit ChatGPT can bring to the realm of fraud detection, claims decisioning, and other processes associated with the policy and claims lifecycles.
ChatGPT: What it is, and what it isn’t
Like any technology, new or established, ChatGPT is a tool. Its ability to create content, find answers, draft emails, or do your kid’s homework based on a set of prompts is genuinely remarkable. Yet, for those impressive capabilities to have an impact they must be part of a larger strategy.
The first step is understanding what generative AI can do for your business and determining what systems must be built around it to reap those benefits. It is also critical to remember that ChatGPT is not perfect. Generative AI may lack the ability to interpret context unique to a specific business, it may be unable to show emotional intelligence in its responses, and because generative AI lacks “common sense,” its responses at times may not make a lot of sense.
Making ChatGPT in insurance work
At Shift, we remain 100% focused on meeting the needs of the global insurance industry. As such, we are in a unique position to help insurers reap the potential of emerging technologies. Innovation can certainly spark experimentation, but operationalization is key to transformation. And as is often the case, execution is considerably easier when you have a trusted partner who can help you understand how new technologies fit into your current plans and strategies as well as pave the way for future transformation. Take just one example, data.
It’s common knowledge that AI is only as good as the data feeding the system. There’s a reason “garbage in, garbage out” has become such a common phrase in the world of information technology and especially AI. For the best results, data silos must be broken down. Data must be cleansed, mapped, and enriched. This is something Shift has been doing on behalf of its customers since the beginning.
Unlocking the power of integration
As previously discussed, ChatGPT’s value is somewhat limited as a “stand-alone” technology solution. While individual ChatGPT tools can be acquired, the true benefit comes from being part of larger systems or solutions intended to solve specific business challenges.
For Shift, that means giving insurers access to generative AI solutions that are essentially “pre-integrated” into their other important core systems such as claims or policy management solutions. Shift can save insurers time, effort and energy in putting these technologies to work for them and on behalf of their policyholders.
Let’s look, for example, at how ChatGPT can be used to augment claims fraud detection via information extraction of corresponding documents. If we visualize a sort of “processing pipe” for insurance documents, the entry point would be optical character recognition (OCR). OCR transforms documents to text to which a ChatGPT question or prompt can be applied.
The result would then pass through what we refer to at Shift as the “Insurance Common Sense” layer to ensure the accuracy and efficacy of the response. We designed this portion of the processing pipe to focus on the challenges specific to the insurance industry and tuned it to the complex scenarios and decisions that drive the industry.
This stage is also critical for continued learning and tuning of the model. Finally, the results are made available to the algorithms to determine if a claim is suspicious and, if so, why, leading to greater accuracy, efficiency and fairness in the claims fraud detection process. This processing pipe approach can also prove useful in other areas such as subrogation detection, underwriting risk detection, and in supporting claims automation strategies and initiatives.
The future of generative AI is here
With so many emerging technologies, it feels as if the time between experimentation to usefulness is nearly infinite. In the case of ChatGPT, that assessment couldn’t be further from the truth, at least for Shift. We’ve already begun working with Microsoft to create a secure private environment for all generative AI services, specifically Microsoft’s Azure OpenAI Service, so these technologies can be brought to bear for the insurance industry's benefit in a safe, scalable and supportable environment. We anticipate these services to be fully operational by the end of Q3 2023.
Unleashing the Potential of ChatGPT: Revolutionizing the Insurance Industry with Advanced AI Decision-Making
It's an exciting time in the world of AI. ChatGPT and other generative AI technologies promise to inject whole new levels of efficiency across many businesses and disciplines. The potential benefit to the insurance industry alone is tremendous. Across the policy and claims life cycle, ChatGPT holds the promise of helping insurance professionals make the best decisions possible on behalf of their companies and their policyholders.
For more information about how Shift can help you adopt generative AI to meet the unique challenges facing the insurance industry – contact us today