Stock market shake-up over AI: will artificial intelligence really replace insurance brokers?

In just a few days, international brokerages such as Willis Towers Watson, Aon and Marsh McLennan lost significant stock market value.

The trigger was an AI use case in the retail sector, where standard products can be quoted via chat.

The narrative immediately followed:

"AI replaces brokers."

But this conclusion falls short.

The actual disruption does not affect the advice itself,
but rather the structure, efficiency and competitiveness within the brokerage.

This article shows why AI is not replacing insurance brokers, where the actual market shift is taking place and why the competition of the future will be decided between brokers with and without AI.

Standardised products in the digital transition

Products such as motor, home contents or personal liability insurance have been digitally comparable for years. Comparison portals have created transparency, and closing processes are standardised.

Artificial intelligence merely accelerates this trend. Today, clients research with ChatGPT, have performance differences explained to them and enter discussions with concrete price expectations.

This changes the preparation, not the role of the broker.

This is because real value creation occurs where risks become complex, individual and liability-relevant.

Advice remains subject to liability

Insurance advice is not a chatbot business. It is subject to documentation obligations, regulatory requirements and liability under law.

Particularly in the case of income protection, life or commercial insurance, it involves health data, long-term financial decisions and individual risk assessments. This is where GDPR, IDD and, increasingly, regulatory requirements for AI systems apply.

A language model is not liable.
A broker is.

Therefore, AI in the brokerage can support, structure and prepare, but it cannot advise autonomously. Human responsibility remains crucial.

The real market shift begins in the back office

The market shift does not take place during the client meeting.
It takes place in the back office.

AI does not replace brokers.
AI accelerates administration.

In daily brokerage work, this means: incoming mail is structured, cases are automatically assigned, documents are checked, deadlines are monitored and replies are prepared. Processes that previously tied up time and attention run structured in the background.

Those who process enquiries in minutes instead of hours react faster, appear more professional and manage more clients with the same team size.

AI does not replace advice. It replaces friction losses.

Not human versus machine, but structure versus stagnation

The development is not "human against machine".
It is: brokers with AI against brokers without AI.

Brokers who systematically integrate artificial intelligence reduce their cost-to-serve, increase their response speed and work in a regulatorily cleaner manner. Brokers without AI remain professionally competent, but operationally slower.

In a market where clients come to meetings well prepared and expect quick feedback, speed becomes a competitive factor.

Conclusion: the course is being set now

AI does not replace insurance brokers.
But it changes the cost structure and efficiency in the market.

The actual disruption does not lie in the advice itself. It lies in the process speed.

Those who use AI intelligently in the brokerage professionalise their business. Those who wait defend rising costs while speed decreases.

The crucial question is therefore not whether AI replaces brokers, but how long a broker can afford to work without it.