Every trend has an impact and corresponding implications.

Avocados are healthy, for example, but avocado demand has caused environmental issues. E-mail, messaging, and tools like Slack help us get answers quickly, but interruptions can harm deeper ideas, better communication, and long stretches of focused thought. Smartwatches track our health and foster good habits, like sleep, but for some personalities, they cause unnecessary stress and obsession. “Everything in moderation” seems to be the universal truth that can help with handling trends.

The insurance industry has its own set of trends, but in our case, it is important to distinguish between technology trends (such as cloud vs. on-premise) and pain point trends (such as increased risks and a growing number of retirements). Technology trends should never be adopted for trend’s sake. “Everybody’s doing it” doesn’t make sense. Tech trends should address business pain points, risk developments, and business opportunities. The hazard, however, is that some tech trends are applied as point solutions instead of holistically integrated solutions. The bottom line is:

  • Trends pose complexity and require us to think ahead and think clearly.
  • Technology and risk trends beg for a response from insurance executives.

In a recent Majesco webinar, I invited three of today’s sought-after InsurTech guides to join me and help paint a clearer picture of today’s trends (such as Gen AI and operational optimization), plus share insights on how insurers should respond. You can tune in to that full webinar, 2024 Trends Reshaping the Insurance Business — Are You Ready?, or keep reading for Part 1 of our conversation, What’s next with GenAI for insurance?

Our panelists are:

Matteo Carbone, Founder and Director, IoT Insurance Observatory

Lisa Wardlaw, Founder and President, 360 Digital Immersion

Bryan Falchuk, Author of The Future of Insurance and President and CEO, PLRB

And myself, Denise Garth, Chief Strategy Officer, Majesco

Let’s begin! For starters, we assess the current state of GenAI adoption in insurance.

Denise Garth

GenAI continues to be the hot topic in the industry with lots of discussion. The rapid adoption of AI-machine learning models, and now the emergence of GenAI have turned data and analytics from a long-term strategy and incremental investment, into a near-term reality and now a must-have investment.

How quickly is GenAI moving from hype to reality? What are you seeing as the key areas of focus for GenAI and the value it brings?

Lisa Wardlaw

There is incredible pressure to be relevant. If you are in the C suite talking to investors and analysts, you have to have a plan for what are you doing with AI and GenAI. However, your data and data strategy and systems foundation may not be up to speed. The one needs to catch up to the other.

The use cases I’m seeing are a little disappointing because they are point solution use cases — fragmented and siloed. I understand creating value and showing traction with AI. I’m concerned about the lack of an integrated strategy. We need to make sure it’s connected to a bigger enterprise strategy, because, as we saw with a lot of digital strategies, incrementalism, without integrated holistic thought, isn’t the best approach.

Bryan Falchuk

Lisa hit on something really important here. We are not talking about an incremental technology. Viewing GenAI that way would be like looking at the Internet 30 years ago and saying, “That technology is just for use on the margins.”

If companies look at the way they operate and try to assess where they can find some value if they “just inject a little GenAI into it,” that’s fine. But ‘fine’ is why we have so many carriers with combined ratios over 100 right now.

GenAI is a completely rethink-your-business-model opportunity. This is one of these moments to say, “What if we holistically rethink this process, and just look at the end product we’re trying to deliver? It’s like a first principles/design thinking exercise, rather than “Let’s use it to fix what we have.”

Matteo Carbone

Let me disrupt the discussion a little bit. I think that it should be necessary to talk more about AI and less about Generative AI. Generative AI is an awesome tool to do some things. But when I sit at the table with insurers that have made a relevant journey already with the different kinds of AI, I hear stories that are pretty disconnected from what we discuss at conferences and read in articles. They have found value and positive ROI using neural networks to make decisions. They have created different models to support underwriting decisions, etc.

So Gen AI is a great productivity tool. Any employee in any sector in the next 12 months will have more of these tools. But it will be more like Excel. You don’t build a competitive advantage because you build a better Excel — you need to be really good at using Excel.

The largest portion of companies that work in the insurance sector will not build their own Generative AI. They will not build their own large language models. They will use many different AI models on their unique insurance data. So, we reconnect with Lisa’s point that it is necessary that your data is well structured, there is a decent governance, and the data is of a decent quality.

Denise Garth

You bring up a great point, Matteo. There’s a difference between Gen AI and what it can do, versus some of the AI machine learning models. Everybody’s placing them in the same bucket.

However, from a Gen AI perspective, when we look at an industry that is going to have a rollover of employees — maybe up to 50% of our employees by 2030 — there’s a real opportunity for operational effectiveness. What’s concerning is that some are saying, “Let’s just kind of take a wait-and-see approach.” But is that too risk-averse for AI machine learning models and Gen AI?

Lisa Wardlaw

I totally agree that we need to focus and that there are “fit for purpose” solutions as we’re reimagining our business — and not all things deserve the same level of octane. But we shouldn’t use AI to be more like RPA band-aids. Certainly, not everything needs to be Gen AI. The fear that I have is that the wait-and-see approach will not be conducive. We have a market that is continually fragmenting a little bit between what buyers need versus what buyers want. Strategic insurance companies, carriers, and innovators will start to ask, “How do we use these capabilities to make insurance products more discernible, more understandable, and more informative? How do we use them to fill a missing gap in our industry by applying the principles of behavioral psychology?”

Bryan Falchuk

Yes. We as an industry have a wait-and-see approach to things and that’s generally been the right answer for a lot of disruptors, like usage-based insurance and the amount of hype around it. With usage-based insurance, you may have some time to build. You should be making some moves. You don’t need to think, “We missed our chance.”

We can’t apply that same timescale to everything that we look at, either. Now, it doesn’t mean you should throw everything away, scrap it, and go work with ChatGPT to redo your entire business. That’s not the answer, either. But you can’t just say, “We’re going to sit back and watch.”

We need to get comfortable with it. We need to find those areas where our people can get more comfortable, where our distribution partners can be comfortable with us using it. But you do not have the right, I think, to just sit by and say, “We’ll see what happens in the next five years and then do something.” You’ll see a lot of things happen, but once you’ve reached the five-year point, it’s too late.

Denise Garth

And it could even be sooner than five years…

So, where does this lead us? What is next with GenAI in 2024?

Matteo Carbone

Applications may be more easily approached because the user interface won’t require a data scientist. Anyone will be able to do their job better using a new technological tool. I would expect that to see a wave of solutions where the interface is generative AI, but it is only a layer that can allow another technical user to interact with other AI models prepared by an expert.

Lisa Wardlaw

I think we’re going to see a much more heightened enterprise approach to AI and AI operating models, and fewer point solutions. Also, within GenAI, we’re going to specifically heighten and hone in on LLM. The pioneers are going to be connecting AI strategy to LLM to LAM, which is the action model. They will be answering the question, “How do we now predict action as a result of our AI strategy and LLM?”

Bryan Falchuk

I’m not sure it’s about, “What’s next?” with GenAI. I still think it’s about, “What’s first?” for most carriers. So, first, there are the initial applications. The things that make the most sense to start to play with are the areas that support the expertise of your people.

Carriers should find areas that, as Matteo mentioned, a non-technical person can get very complex information brought together at the right moment in a way that they can request and digest. They can have data presented right within their workflow. That’s what you’ll start to see adopted this year because it doesn’t take epic shifts of your entire operating model. People will begin to get comfortable with it.

This approach supports the fact that we have huge retirement rates in this industry, as well as people very separated with remote work, and the result is that we aren’t receiving adequate knowledge transfer. These tools will be super valuable.

Denise Garth

Excellent! I would like to thank our panel for their insights. In our next installment, we’ll discuss operational optimization. Which technologies can improve operations, products, and profitability?

What’s next for your organization on AI and GenAI?

Does your organization need to solidify and launch its AI and GenAI strategy? Are you interested in how your organization can take a holistic approach to AI and GenAI while growing comfortable with all that it can do? Begin with Majesco’s recent webinar, Revolutionizing Insurance: The Power of Majesco Copilot & Microsoft Generative AI, on how we have embedded AI models and Gen AI into our solutions to leverage their power. or contact us for more information. Whether your focus is customer service, operations, underwriting, loss control, billing, or claims, you’ll find that Majesco Copilot answers your needs for simplification, automation, and timely information.

By Denise Garth