AI in the Swiss Insurance Market: What Insurers Can Learn from the Results

What Really Influences Acceptance

In the first part of this article, we showed how widespread artificial intelligence already is in everyday life, how much trust people place in artificial intelligence when it comes to insurance matters, and in which areas customers accept its use. It became clear that while there is a general openness to AI, this openness diminishes as AI gets closer to individual decisions, benefits, or the assessment of specific cases.

Now let’s take a closer look at the differences among customer groups. Age, income, and—above all—personal experience with AI influence how open people are to such applications. For insurers, therefore, the question is not so much whether they should use AI, but rather where, how, and among which target groups such use will be well-received.

Even experienced AI users set limits

While extensive experience with AI generally increases openness, it does not eliminate these barriers. Among those who never use AI, 65 percent do not accept any of the application areas surveyed. Among daily AI users, this percentage drops to 12 percent.

Nevertheless, they, too, make a clear distinction between different applications. For example, 45 percent of daily users accept the use of AI in customer service. When it comes to processing insurance claims, however, that figure drops to just 24 percent. Experience with AI thus fosters openness, but not unconditional acceptance.

Age and income make a difference

In addition to individual experiences with AI, there are also significant differences among various customer groups.

As people get older, both the use of and trust in AI decline. However, this does not mean that older customers are fundamentally opposed to AI. Among those over 60, for example, 40 percent accept AI for fraud detection. In customer service, on the other hand, that figure is only 23 percent.

There are also striking differences in terms of income. In households with a monthly income of more than CHF 12,000, 51 percent accept the use of AI for fraud detection and 39 percent for processing applications. In contrast, among households with a monthly income of up to CHF 4,000, 43 percent say they do not accept AI in any of the areas surveyed.

For initial pilot projects, AI-savvy target audiences can therefore be particularly interesting. However, anyone looking to scale up later should not make the mistake of assuming that this will apply to the entire market.

What does this mean for insurers?

Three key guidelines for the use of artificial intelligence can be derived from the results.

1. Start where the benefits are immediately apparent

Fraud detection, customer service, and quote generation are currently the areas with the highest level of acceptance. These are precisely the areas where it makes sense to roll out initial applications and enable positive experiences with AI.

2. Use AI as a co-pilot rather than as the sole decision-maker

For applications, recommendations, and similar uses, results should remain transparent. At the same time, it should be clear to customers where human oversight, correction, or confirmation is possible.

3. Highlighting Human Responsibility in Sensitive Decisions

Customers tend to be particularly cautious in situations where AI evaluates individual cases, such as when reviewing insurance claims or in highly personalized applications. In these areas, it is likely particularly important to clearly demonstrate where human responsibility remains.

The key question for insurers is therefore not so much:

Where can we automate as much as possible?

Rather:

Where does AI create tangible value for customers without compromising trust and control?

That is likely where the most sensible starting point for a long-term, sustainable AI strategy lies.

About the Study

The ValueQuest study “AI in the Swiss Insurance Market” examines the use of artificial intelligence, trust in AI, and the acceptance of various applications in the insurance sector.

A total of 2,080 people between the ages of 18 and 67 in German-speaking and French-speaking Switzerland were surveyed in April and May 2026. The results are descriptive and highlight differences and patterns among various target groups. No causal relationships can be inferred from them.

The analysis was conducted as part of ValueQuest's Moment of Truth Tracking (MoTT).

Would you like to learn more about how Swiss insurance customers perceive AI and which areas of application offer particular potential? Contact us for more information about the study and Moment of Truth Tracking.