Who’s Leading Insurance AI? We Built the Sapiens Insurance AI Maturity Index to Find Out

I’ve spent a lot of time this year talking with insurance leaders about AI. Everyone has a project underway, a pilot in motion, and a team studying the question. What struck me, though, was how little anyone could tell me about where they stood relative to their peers. Everyone had an anecdote. Nobody had a benchmark. 

That gap is why we built the Sapiens Insurance AI Maturity Index, a global study of 475 senior insurance professionals across the UK & Ireland, the Nordics, Benelux, South Africa, and North America, spanning both life and P&C carriers of every size, from specialist firms with fewer than 1,000 employees to global groups with more than 20,000. 

This Index answers a simple question: how mature is the industry’s use of AI, really, and who’s pulling ahead? The study was unveiled at the Sapiens Ignite Summit, and I want to weigh in on the highlights.  

What the Index measures 

The Index sorts insurers into four maturity categories: Reactive, Enabled, Operational, and Autonomous. The scoring is built on four areas that determine whether an AI strategy sticks or stalls: cloud adoption, master data strategy, AI governance policy, and HR strategy. As we close out 2026 and set goals for the coming year, we focused on the fundamentals: clean data, sound governance, and a workforce that’s been trained to use the tools. 

With more than 40 years of experience in the industry, we know that without the right tools and the proper framework, AI maturity is impossible. We also know that “maturity” is a moving target, which is why this Index is not a one-off snapshot. It’s something we plan to do every year, so the industry can track its own progress. 

What we found 

The finding that struck me most is the industry’s optimism: 67% of respondents believe AI in insurance will be fully autonomous within two to three years, or sooner. That is not a distant, theoretical horizon. That’s the planning window these businesses are already operating inside. 

At the same time, there is also hard-edged realism. Only a quarter of insurers, the group we call Autonomous Leaders, are where they need to be to compete in that world. These are the companies on the cloud, often with SaaS, with master data and HR strategies established, and governance policies in place. As we all know, the cloud is essential because for AI to be effective, it must draw data from a common pool, across the entire organisation.  

Depending on your point of view, 25% is either a low figure or a high one. Personally, I look at it as a positive development, like the 28% of respondents who are Operational, with projects in motion. As for the remaining nearly half of the industry, they are Enabled or Reactive, which is another way of staying they are early in cloud adoption, early in data strategy, with governance and workforce planning still catching up. 

The gap between automation, AI, and agentic AI is just as telling. On average, 40% of insurers say they’re using automation and 42% say they’re using AI, but only 24% have moved into agentic AI: systems that plan, decide, and complete multi-step tasks independently. Interestingly, 87% of respondents say even though agentic AI is important to their future, only 16% have adopted it. That’s the widest intention-to-action gap in the entire study, and it tells me most of the industry knows where it needs to go but hasn’t yet built the muscle to get there. 

There’s a real payoff for the insurers who have made the leap. AI is delivering measurable gains in turnaround times, customer experience, error rates, and manual handling costs, with 77% of insurers saying they now identify and deploy products more quickly because of it, saving an average of three months on time-to-market.  

I was also encouraged, and more than a little reassured, by what we learned about trust. Confidence in AI-driven decisions is high: 86% of users are very or somewhat confident in the calls their AI and agentic tools make. But insurers aren’t handing over the keys blindly. On average, 53% of agentic AI users still have a human checking the technology’s work, and only 15% say they need to override AI outputs regularly. That combination of high confidence and continued oversight is exactly the posture I’d want to see at this stage: trust that’s earned through verification, not assumed. 

Why this matters right now 

The leaders and the laggards are not separated by resources or geography as much as you’d think. Yes, larger, better-resourced organisations tend to score higher, and regions like Benelux and the Nordics are ahead of the UK. But the real differentiator we found is people. Autonomous Leaders are investing heavily in upskilling their workforce (71%, versus 53% globally) and moving faster to bring in AI-savvy talent (57% versus 38% globally) to transform their organisations. 

That’s the message I’d leave you with. The industry has largely made up its mind that autonomous AI is coming soon. The question this index is really asking is whether your organisation is building the cloud infrastructure, the data foundation, the governance, and the talent base now, while there’s still time to move from Reactive or Enabled into the Operational and Autonomous tier. Two to three years sounds like a long runway. In my experience, it isn’t. 

As you draw up your goals for 2027, we hope that the Sapiens Insurance AI Maturity Index gives you the context to plan for where the industry is heading. Everyone deserves a real benchmark; we hope it serves you well and becomes an invaluable resource every year when the latest findings are shared.