Our AI journey so far: Tjerrie Smit discusses story behind NN’s use of AI
NN Group has been developing its AI capabilities since 2015, moving from experimentation to large scale applications across its businesses. Today, the focus is shifting towards agentic AI: systems that can take actions within business processes, rather than simply answer questions.
In a recent interview with podcast, AI at Work Live, NN Group’s Chief Analytics Officer, Tjerrie Smit, shared his perspective on what this next phase of AI means for NN and the wider insurance industry.
For insurers, the shift to agentic AI creates opportunities to make routine work faster and more consistent, while carefully considering cost, governance, and human oversight. As of 2026, agents are now carrying out complete sequences of work across parts of NN Group’s businesses.
From assistance to action
‘AI is not just a chatbot anymore; it can perform tasks for you or do something on your behalf,’ says Smit. One example is a straightforward glass-damage claim: in the past, a typical claim required fourteen different checks. AI agents can now carry out all those fourteen tasks, allowing employees to focus more on exceptions, judgement, and accountability.
Choosing the right model and keeping people in control
Scaling AI also changes the economics of technology. Usage-based token costs mean organisations need to understand not only where AI is deployed, but how often it is used and what value it creates. The most powerful model is not automatically the best fit for a task. Smaller, specialised language models may offer a more efficient option for specific applications, with potential benefits for cost, energy use, and governance. As Smit puts it: ‘A sports car is fun, but you do not always need a sports car.’
Human oversight remains central, especially when decisions have major financial, legal, and emotional consequences. Routine steps can be automated, but accountability does not transfer to the technology. ‘For a high-impact claim, you ultimately want a person to review it,’ Smit says. This risk-based approach keeps people in the loop where judgement matters the most.
The importance of learning new skills continuously
NN Group’s experience points to a practical principle: start with the business problem, not the technology. Leaders need to define the outcome, select the appropriate model, and set clear boundaries for human review. Because ultimately, it is leaders and their teams who remain accountable for the output AI produces.
At the same time, managers and employees will need to keep adapting as AI changes how work is organised. The most durable strategy is continuous learning: understanding what AI can do, where its limits lie, and how people can work with it safely and effectively. NN Group helps colleagues to learn these new skills with an extensive Data Literacy programme.
Want to hear more from Tjerrie Smit?
You can listen to the Dutch podcast episode from AI at Work Live (S2E1, from 30:32) via Spotify.