McKinsey’s insurance practice just hosted more than 50 chief claims officers and other senior insurance claims professionals from 19 countries at our latest European Claims Forum, in Vienna. The discussions came at an important moment for the industry: Claims inflation and complexity continue to pressure insurers, customer expectations are rising, and a potentially softening pricing environment is putting renewed emphasis on operational performance.
While much of the discussion focused, inevitably, on how AI is changing the claims business, the technology itself is just the starting point. The key is how to use it in a smart and targeted way to truly boost performance. What stood out was that the opportunity goes well beyond automating individual tasks. Agentic AI creates the possibility of fundamentally rethinking the claims journey, although the right approach will differ between high-volume personal-lines claims and complex specialty claims.
Following are three takeaways from a lively day of discussion.
1. Fundamentals still matter
Yes, the technological advances are exciting and important, but it is much easier to extract value from AI if you have the fundamentals in place. That calls for strong claims expertise, disciplined processes, effective supplier and partner management, and the ability to consistently make sound claims decisions. AI can amplify the capabilities of a well-run claims organization, but it cannot compensate for weak processes or a lack of domain expertise. In fact, deep claims knowledge becomes even more important when it is embedded in AI-driven processes and decision-making.
Using our proprietary Finalta claims database for benchmarking, we can see that most of the top players have started to use AI for aspects of their claims handling, but that this is not universally the case. Some of the best performers are behind on AI implementation but make up for it with excellent processes. While that still works today, it may not be enough for tomorrow. Still, the other way around—layering AI onto underperforming processes—is not an effective solution. If you just have the technology, you still need expertise, a clear sense of what matters, and the ability to pass this on to the AI models.
The Finalta benchmark analyzes about 52 million motor and home claims across Europe over three years. Looking at operational metrics, top-quartile carriers settle claims 16 percent faster than the average carrier. In indemnity spend, where the gap between the top-quartile carriers and the average narrowed sharply after 2020, leaders started to pull away again in 2025.
How much of this performance gap can be attributed to AI is not yet clear. Finalta’s benchmarking shows that strong operational fundamentals remain critical, with some high-performing insurers achieving excellent results despite relatively limited AI adoption. At the same time, as AI capabilities mature and become more widely embedded in claims operations, they could increasingly differentiate performance—particularly by helping insurers improve claims quality, productivity, and customer experience simultaneously and without the need for trade-offs, as has been the case historically.
2. Being bold with implementation
The arrival of AI sparked a flurry of experimentation, as claims organizations set up dozens of pilot programs to identify and understand the best use cases. At our Claims Forum, we heard from many how that phase has been giving way to more in-depth implementation with the goal of scaling up usage. In practice, that doesn’t mean trying to do everything at once but rather going deep in selected areas.
Two case studies presented at the forum illustrate the power of this approach. One Dutch insurer illustrated how a focused approach can deliver significant results. Rather than deploying AI to solve individual problems, the company set out to transform entire claims domains, starting with high-volume retail claims and then expanding into more complex areas. The insurer reported substantial improvements in claims-handling productivity and processing speed. Importantly, achieving these results required not just new technology but also a fundamental redesign and deep focus in selective areas.
The second example illustrates a very different application of AI. A specialty insurer processes complex claims involving extensive documentation, multiple stakeholders, and difficult coverage questions. For such claims, AI can help experienced adjusters systematically review information, identify potential inconsistencies, and assess issues that might otherwise be overlooked. Rather than automating the entire claims journey, the insurer used the technology to help improve the quality and consistency of decisions by enabling access to relevant information and secondary analysis. Experienced claims professionals retained responsibility for the final judgment.
3. Addressing the leadership challenge
Ultimately, the biggest challenge may not be technological but organizational. Claims leaders need to set a bold ambition for how AI can transform their business, bring experienced claims professionals into the redesign of processes, and ensure that productivity improvements translate into tangible business outcomes. This requires rethinking roles, skills, and ways of working—not simply introducing new technology.
At the same time, the future claims organization will increasingly involve humans and AI working together, with different levels of automation depending on the complexity and nature of the claim. For some high-volume claims, AI may orchestrate much of the process, while in more complex cases, experienced professionals will continue to play a central role in judgment, negotiation, and customer interaction. The leadership challenge is to design this collaboration effectively and build the organizational capabilities to make it work.
The discussions at the Claims Forum underscored how claims is at an inflection point. Agentic AI offers the potential to do much more than improve individual processes; it allows leaders to fundamentally rethink the claims operating model itself. The question is whether insurers will use this technology simply to optimize how claims are managed today or seize the opportunity to reinvent the journey end to end. Insurers that combine strong claims fundamentals with bold ambition and disciplined implementation will be best positioned to translate technological progress into lasting competitive advantage.