The AI race: Who is leading, who is following, who is falling behind?

Readingtime: 4 min

Artificial intelligence (AI) will profoundly change the insurance industry. The business model—risk protection and diversification—will remain, but the operational models will undergo fundamental change. With the availability of this technology, change is now unstoppable. Introducing AI is not an option, it is a mandate.

The difference between insurers with an AI-driven operating model and those that continue to work traditionally will be so drastic that waiting would be fatal. As in the long-tail business, the pain only comes when it is too late. At the same time, the industry is struggling to address the issue with the necessary consistency – for understandable reasons.

The central question is therefore: How do you proceed? And closely related to this: What is actually different this time? This month, I will focus on the second question. Because if you don't know the terrain, you can't plan a route.

1. AI affects the entire company, not just IT

Of course, chief information officers, chief technical officers, and chief digital officers are at the center of many initiatives. But the introduction of AI is not a technical issue – it is strategic. It changes how products are created, how risks are assessed, and how customers are served. The board of directors must declare AI a top priority and promote it as such in all departments.

2. Everyone understands something different by “AI”

The spectrum is enormous: it ranges from rule-based chatbots to external language models (large language models, LLMs) to autonomous agent systems in complex cloud architectures. This diversity makes penetration difficult, and without a common vision, transformation remains piecemeal.

3. It's new, and it's different

The introduction of AI is not comparable to a software rollout. Until now, familiar processes have been “electrified.” AI, however, requires a rethinking of processes – comparable to the introduction of shipping containers, which not only made transport more efficient but also restructured global supply chains.

4. The pace overwhelms traditional control logic

Technological development is progressing exponentially. Classic strategy project roadmaps are outdated by the time they are approved. Anyone planning today has to make adjustments during the pilot phase and should think about “production” at the same time. Those who get stuck in the pilot loop gain experience, but don't save a cent.

5. The organization is not prepared

Insurers are masters of structure – with a high degree of maturity, efficiency, and process orientation. But that is precisely what makes them resistant to disruption. Most companies have rarely experienced fundamental transformations. The first real AI deployment is a big one. And it affects the organization, leadership, and people alike. Those who are good at optimization find it difficult to rethink – and vice versa.

6. Regulation is ahead of the curve

The EU's AI Act is not yet established, but its shadow looms large. A regulatory framework is emerging before many insurers have even started using productive AI. It's like putting up highway signs before the first car has been delivered. Well-intentioned, but challenging for a hyper-regulated industry.

7. The good news

Established insurers have two starting advantages: data and experience. Insurtechs have neither. Those who can launch AI with real data and business cases will see results faster. Insurtechs must first win customers before their technology can take effect. Established companies have experienced technical experts. This is essential for applying AI where it is most effective. An email pilot is nice, but it requires no strategy, no transformation, and the impact on the loss and cost ratio will be slow to materialize.

Established insurers are starting this new race from pole position, while insurtechs have the better engine. At Digital Tuesday on August 5, I will show you which strategies you can use to take the lead – and stay there. Because pole position is no longer relevant once the race has started.

German Version: Das KI-Rennen: Wer führt, wer folgt, wer fällt zurück?

Read more: Das KI-Rennen: Wer führt, wer folgt, wer fällt zurück?

Fitting to the topic: