AI: Building or buying? What insurers are doing

At this stage of the insurance industry’s journey in artificial intelligence, insurers are increasingly opting to build versus buy solutions, according to several experts who spoke with InsuranceNewsNet.

“Because so many people have been burned with buying, they’re leaning toward building. Last year, everything was, ‘Oh, I’ve got to do something in AI.’ This year, it’s much clearer. We’re seeing AI-led development and specific projects popping up where we are doing something and it has value,” said Paul Tiede, partner, insurance practice with Capco US.

He noted that previously, a company would typically buy an admin system and then stick with it for 20 to 30 years. However, he believes AI is now “blowing this up” because it “takes the cost and complexity out of implementing these solutions.”

“I’ve been in the industry for about 25 years now and I think, with AI’s acceleration in the last few years, the build versus buy debate has now increased more than I’ve probably seen in the last 10 to 15 years,” said Chris Raimondo, EY’s global and U.S. consulting leader, insurance sector.

In his view, while “there was a heavy index toward buy” for the past 10 to 15 years, “AI has really changed that in the last couple of years, particularly in the last 12 months as the AI technologies and the frontier labs have really matured and started to make more traction in the industry itself.”

However, Ben Schwartz, global product owner of AI platforms at Manulife, believes many are implementing hybrid solutions — buying some solutions and building others.

“I think most people are not willing to be dependent on a single technology or a single vendor, plus there’s the clear speed of change. And so, in most people’s case, I think they are building some things and buying some things,” Schwartz said.

Choosing build over buy

According to industry leaders, insurers are now choosing to build instead of buy due to:

» Long-term cost and sustainability in a rapidly changing environment

» Opportunities to differentiate from competitors

“It doesn’t necessarily make sense right now to lock into three- and five-year contracts with vendors the way we previously did around certain technologies, just because we know the technology will change and processes are changing so rapidly,” said Michael Corrigan, chief information officer at World Insurance Associates.

“As an annuity insurer, I can now stand up a policy admin system for about the same amount of money as what I can buy a solution for, building it custom, using AI system development. That fundamentally changes the calculus because I’m not locked into this 20-year relationship where I pay a fortune of money for support. I have a custom app that I can control and get exactly what I want for about the same cost,” Tiede said.

For Manulife, Schwartz said the decision can also be a matter of simply not finding existing market solutions for specific use cases.

“Our approach is to build first and buy later if the capabilities that we need indeed mature and emerge,” he said. 

Hybrid options are viable

Schwartz also said AI solutions can be a unique combination of build and buy that fits more specific needs and drives competitive advantage.

“I think there is a strategic hybrid decision where we are buying, but we are deploying it by building something to glue together both the technologies themselves and the way the teams work in those technologies to make sure that we build in a way that is more creative and more efficient,” he said.

In Corrigan’s experience, companies buy solutions for “common capabilities” such as document processing or automated incident response. But then they build solutions for “something niche or bespoke for our business” that can foster competitive advantage through differentiation.

“We’re seeing a little of that hybrid functionality be a bit of a sweet spot. You can get some of the best of both worlds where you can bring in a proven platform, integrate it with your data models and your process flows and then you have the innovation that comes from a vendor-provided solution while getting the tailor-made model that applies to how your business operates,” Corrigan said.

Workflow vs. systems

Avitesh Kesharwani, senior principal consultant at Genpact and AI researcher at the University of North Carolina at Charlotte, said the build versus buy decision isn’t about technology but about workflow expertise and intellectual property.

“If the value is coming from commodity capability, just go ahead and buy it. But if the value is coming from the unique workflow of your enterprise, of your particular enterprise or data or expertise, that’s usually where building becomes worthwhile,” Kesharwani said.

“I think the subtle point that is sometimes missed in the buy versus build decision is that it is really not the technologies themselves that solve the problem; it is how clever, efficient and well-executed we are when we use those technologies,” Schwartz said.

“Think of it more as maybe not building systems. Today, where we’re at in the build life cycle, we’re really building workflow and capability that sits above the core transactional system,” Raimondo said. “I don’t think we’re there at the point yet where those systems are being fully rebuilt. There’s certainly some discussion around it, and I think we have some time to go there.” 

Rayne Morgan is a journalist, copywriter, and editor with over 10 years' combined experience in digital content and print media. You can reach her at [email protected].