AI alone will not create a lasting competitive advantage in insurance

By discovermarket  |  Updated 11 July 2026  |  5-minute read

AI alone will not create a lasting competitive advantage in insurance

AI will matter in insurance. Distribution data will decide who wins.

Over the past months, AI has become part of almost every serious conversation I have with insurers, brokers, and distribution partners. The starting point is not always the same. Some companies are looking at underwriting support. Others are focused on claims, servicing, fraud detection, sales conversion, or customer engagement. But the direction is clear: everyone wants to use AI to reach more customers, improve profitability, and create better customer experiences.

That focus is understandable. Insurance still has a heavy operating model. Underwriting, claims, servicing, fraud checks, compliance, and distribution support all carry a lot of manual work. If AI can reduce part of that burden and help teams make better decisions faster, the business case is clear.

My concern is that many insurers are starting to treat AI itself as the source of future differentiation. I do not think that will hold for long. Most insurers will eventually have access to similar AI capabilities. They will work with many of the same providers, automate similar workflows, and use comparable external data sources. AI will become part of the standard technology stack of the industry.

That does not make AI less important. It means the real advantage will depend on what insurers can do with it, and especially what data and customer context they can connect it to.

In insurance, a lot of that context does not sit inside the insurer. It sits inside distribution. It sits with banks, telcos, retailers, travel platforms, mobility platforms, employers, subscription businesses, and other digital ecosystems where customers already buy, book, move, pay, subscribe, finance, search, or interact.

This matters because a model is only as useful as the context it can read. If an insurer mainly works with internal policy data, past claims, and broad customer segments, AI may improve internal efficiency. But it will struggle to make insurance truly relevant at the moment the customer may need it.

For P&C products, this can be very direct. A customer buys a device, books a trip, rents a car, leases a home, buys an event ticket, or ships an item. Each of these moments can create a protection need. If insurance is offered much later, through a generic campaign, a large part of the relevance is already lost.

For Life and Health products, the signals are different, but the logic is similar. Financial behavior, family changes, employment context, lifestyle indicators, digital engagement, and other consent-based data points can help insurers understand when a customer may be entering a different stage of need. In some cases, broader signals such as social media data points may also help build context, provided consent, regulation, and data governance are properly respected.

This is where AI recommendation and contextualization become important. The point is not to push more products into more channels. The point is to understand which product is relevant, for which customer, through which partner, and at which moment.

Embedded insurance gives insurers proximity to these distribution moments. AI can help interpret those moments, recommend the right product, support pricing, improve servicing, and reduce friction in the customer journey. The value comes from combining access, data, timing, and execution.

This combination can change the economics of insurance distribution. Better context can reduce wasted acquisition spend. Better timing can improve conversion. Better product matching can increase relevance. Better data can also support more disciplined underwriting and servicing decisions.

This is important because many insurers still distribute insurance in a way that is too broad. Products are pushed through campaigns, segments, and static journeys. Customers are often asked to provide information that another part of the ecosystem already knows. The result is lower relevance, more friction, and a weaker experience.

AI can help solve part of this problem when it is connected to the right data and the right distribution environment. Without that connection, insurers may become more efficient internally while still remaining too far away from the customer moments that create growth.

That connection requires distribution partnerships, configurable products, controlled data orchestration, and the ability to connect insurance systems to external ecosystems without rebuilding everything for each partner. This is one of the main operating challenges insurers will need to solve if they want AI to create real commercial impact, and not only internal productivity gains.

At discovermarket, this is the role we play. We help insurers connect existing core systems to embedded distribution environments, so they can launch, manage, and scale digital insurance propositions without turning every opportunity into a long transformation project.

I have no doubt that AI will have an important role in the future of insurance. The question is whether AI alone will create lasting differentiation once similar capabilities become available across the market. I do not believe it will.

The insurers that build stronger access to digital distribution will have something more difficult to copy: the customer context around the risk itself. That is where I believe the next competitive advantage in insurance will come from. Less from the AI narrative itself, and more from being close enough to where customer behavior actually happens.