When AI agents reshape insurance distribution

Historically, the distribution of insurance products has been based on a mosaic of channels: salaried insurer networks, independent brokers, tied agents, online comparison services... Each has found its place, with its own strengths: close relationships, commercial clout, technical specialization or digital excellence.

But the arrival of AI agents - artificial intelligences capable of understanding, dialoguing, researching, comparing and acting - is changing the rules of the game. For the first time, a new technological player can potentially aggregate several functions hitherto divided between specialized channels: advice, comparison, personalization, underwriting.

What if tomorrow's most efficient intermediary wasn't a human or a website, but a conversational agent integrated into an application, a super-app or a search engine?

Against this backdrop of potential market reshaping, this article examines two forms of distribution that are particularly concerned: online comparators and brokerage, both of which could see their value captured by artificial intelligence.

AI agents: the new frontier of artificial intelligence in insurance

For as long as insurance has existed, it has been an industry of data and algorithms: actuarial science is proof of this. The real novelty today lies in the speed, computing power and interaction offered by modern AI, driven by the cloud, massive data and generative models.

Since the invention of artificial intelligence in the wake of the first statistical models, there have been three main stages in the maturity of this technology, and the associated impacts on business have evolved:

AI detective data analysis, anomaly detection, predictive modeling (fraud, health risk, claims).

Generative AI content production, natural dialogue with customers, simulation of complex scenarios.

Agentic AI or AI agent : autonomous agents capable of planning, executing and optimizing tasks, interacting with other systems and databases.

Exponential impact on the insurance value chain

Long confined to support functions (HR, finance, compliance...), AI is now inviting itself into the heart of the insurance business. From product design to distribution, from customer relations to claims management, the entire value chain is being transformed.

The irruption of AI agents in retail: a new turning point.

These artificial intelligences no longer simply analyze data or generate content. They act. They plan, execute, compare and personalize. Simply put: they distribute.

Examples abound in other sectors:

On the journey, platforms such as Mindtrip organize complete stays (flights, accommodation, activities) in a matter of seconds, thanks to intelligent agents.

In e-commerce, from personal assistants guide customers from product selection to logistics.

In HR, some tools already control sorting applications and scheduling interviews.

The insurance industry will not escape this wave. Distributionwhether direct or through intermediaries, will undoubtedly be the first domino to fall under the influence of these new digital intermediaries.

And if all forms of distribution are concerned, two models are particularly at risk the brokeragethe historic pillar of the customer relationship, and online comparatorsspearheading the digitalization of the sector.

Reinventing the brokerage market with AI agents

A fragmented sector

In France, the brokerage business represents around 26 % of the individual health insurance market and more 50 % in commercial general insurance ( France Assureurs, Rapport 2023 sur la distribution en assurance santé et IARD).. This sector is structured around four main families of players.

Local brokers - more 18,000 in France (ORIAS 2023 - Registre unique des intermédiaires) - are the historic foundation of the brokerage business. Their added value is based on human relations, proximity and personalized support for policyholders, whether individuals or small businesses.

Wholesale brokers occupy an intermediary position between insurers and local brokers. They design and distribute products, often innovative, which they make available to a local network, while providing services (digital platforms, management tools, training).

Major international brokers are distinguished by their sector expertise and their ability to manage complex risks, particularly for large corporations and international markets. Their added value goes beyond simple distribution: they offer advice on risk management, access to reinsurance and customized services.

Finally, online comparators have already captured a growing share of the market thanks to the speed and transparency of their services. They alone account for nearly 42 % of French insurtechs (Propulse by Crédit Agricole, Survey of insurance brokers and agents in France, 2023)This is proof of the segment's vitality.

Faced with this, the AI agent acts as a universal intermediate. It combines extreme customization processing and vision transversee dmarket. It is therefore likely to reshuffle the cards.

From consequences real for brokersvariables according to the nature of their business:

ACTOR

FORCES

WEAKNESSES

RISKS

Major brokers

Investment capacity

Value-added services for their customers (multi-risk management, international...)

Negotiating power

Weaker customer proximity

Emergence of AI agents with an international view of risks and markets

Wholesale brokers

Negotiating power

Critical size for complementary services (to customers / local brokers)

Model exposed to the health of local brokers

Need to rapidly integrate interfacing with AI agents

Pressure on margins if strong emergence of new intermediaries (AI agents)

Local brokers

Historical proximity to customers

Strong territorial roots and knowledge of the local fabric

Weaker capacity for innovation: financial volume, insufficient size...

Weak digitalization

Dependence on wholesale brokers

Absorption of their role by AI agents ("digital hyper proximity")

Accelerating market concentration

Online comparators

Transparency and simplicity

Current positioning as a digital intermediary

Added value sometimes limited to comparison

Dependence on web referencing

Little differentiation between market players

Risk of marginalization if customers go through AI agents

Obligation to transform into an AI agent platform

In this landscape, AI agents risk accelerating market concentration and imposing a profound redefinition of the role of each player.

These developments leave little room for complacency. To stay in the race, insurers and intermediaries alike must act now.

Act now: key levers for insurers and brokers

A fragmented sector

The coming transformations leave no time to linger: players need to prepare their systems and models now. Several key levers can be activated to kick-start this transformation.

Structuring data. This stage involves making customer databases, offers and services more reliable. Without clear data, there can be no effective distribution by an AI agent, and no fluid interaction with customer agents.

Modernizing information systems. Cloud, APIs and modular architecture (specialized bricks) are no longer strategic choices, but the foundation of economic sustainability, ensuring new distribution capacity.

Identify ROI areas and test on a small scale. The aim is to target distribution processes (prospecting, underwriting, referral, etc.) with the greatest business impact before launching POCs, to avoid spreading resources too thinly.

Defining an AI-based business model. AI doesn't just change technology, it imposes an organizational transformation including :
Visit process reinvention (distribution, management, customer relations).
L'changing professions (agent supervision, data management).
Visit governance (AI Act compliance, human supervision).

In the United States, for example, job offers in the insurance industry mentioning an AI skill have risen from +16,15 % between 2023 and 2024 (Artificial Intelligence Index Report 2025, Standford University)proof that the movement is already underway.

Three actions to get the ball rolling

The technical and technological challenge is immense, but the transformation must not be thought of as a big bang. Success will come from a progressive approach combining strategy, technology and change management. Three actions, which should not be approached sequentially at the risk of blocking any initiative, appear to be decisive:

Deploying AI agents The aim is to assess the value chain, identify concrete use cases, prioritize those with the highest ROI (distribution, underwriting, management), and then experiment with and scale up the projects identified.

Building a solid technological base. This stage involves aligning data architecture and information systems with market standards (cloud, API, microservices), while integrating regulatory and security constraints specific to the insurance industry.

Supporting teams. Finally, we need to define clear governance, provide training in the new professions (AI agent supervisors, data stewardship, value management), and encourage gradual appropriation by employees and distribution networks.

AI is not (just) about technology, but about strategy and transformation. What's at stake here goes beyond mere automation: it's about the place each player wishes to occupy tomorrow.

Conclusion

Insurance has always been an industry of data and statistical models. AI does not represent a total breakthrough, but rather a decisive acceleration. After detective and generative AI, the AI agent ushers in a new era: that of a universal digital intermediary, capable of redefining insurance distribution and dramatically shaking up the brokerage market.

As with the digital transformation of the 2010s, the players who anticipate, test, modernize and support will be ahead of the game. The others will see their role diminish in favor of digital intermediaries capable of extreme personalization and real-time interaction.

The AI agent does not mean the end of brokerage. Nor does it mean the end of comparators. But it is the start of a strategic repositioning in which the players who know how to transform themselves - technically, commercially and humanely - will come out on top.

Conversely, those who remain stuck on historical models could see their value captured, their role marginalized, and their proximity challenged by automated hyper-distribution.

An article by

Antoine Evennou

Principal

Camille Dugelay

Manager

Félix Humbaire

Partner

Do you have a project? Have a question?

Our teams are ready to listen to your needs and work with you to develop tailor-made solutions.