AI was the dominant investment theme in insurtech at the start of the year. A record $820 million, or 97%, of disclosed insurtech funding in the first half of the year went to AI-focused companies, according to the Simon-Kuchers Insurtech Tracker. Not only did AI firms attract nearly all available capital, but they also raised substantially larger rounds than their non-AI peers.
As we reach the midpoint of 2026, the conversation is evolving.
If the first half of the year was defined by a race to bring AI-powered products to market, the second half will be defined by commercial discipline and pricing realities.
First half lessons
The first half of the year provided some cautionary lessons.
There is a realization that AI functionality alone is not a defensible moat, and B2B insurtech winners will have to own critical workflows, proprietary data or distribution channels, and will deliver measurable outcomes.
Regulators and courts are also holding enterprises accountable for AI-driven decisions, regardless of whether the tools were built in-house or externally sourced. In a recent incident involving a policyholder whose coverage was cancelled during a review period, the Pennsylvania Attorney General found "alleged unfair or confusing auto insurance cancellations" related to an AI tool at the carrier. The Pennsylvania incident and the
Insurtech solutions that can provide explainability, model monitoring and documentation around AI solutions will be better positioned to win enterprise buyers. Vendors that cannot will face longer sales cycles and more limited deployments.
Second half predictions
The pressure to launch something, anything, related to AI is intense.
However, the winners over the next 12 months will not necessarily be the companies that launch the most AI initiatives. They are companies that solve painful distribution and underwriting problems for incumbents or sit inside high-value workflows. The focus is already shifting from bringing AI products to market for the sake of participation to identifying areas where AI can create material business impact.
The market will move away from AI tools or point solutions toward ownership of end-to-end workflows.
AI tools that address narrow tasks like summarizing a submission, drafting a claims note or extracting data from a document are easy to replicate. They are also increasingly difficult to defend as delivering compelling business outcomes.
By comparison, delivering AI infrastructure at the operational layer requires complex coordination across departments and functions, and can directly impact expense ratios, cycle times, risk selection and incumbent economics. End-to-end workflows will be the next battleground.
Outcome-based pricing will be gradual and challenging
Insurtech providers are exploring outcome-based pricing models that align fees with value created such as the number of claims processed, fraud incidents prevented or policies bound.
Conceptually this makes perfect sense. The challenge is that most insurtech firms are not set up for outcomes-based pricing. Data on how their solution reduces handling costs, shortens cycle times, impacts retention or leads to premium growth are often inside the client's organization and not readily accessible. Additionally, the telemetry that exists to support outcomes-based pricing may create more friction than progress depending on what they are doing.
The destination appears clear, but the path will likely be gradual and challenging. The industry will move through a series of intermediate pricing models before arriving at a true outcome-based pricing model.
Ultimately, insurtech firms with a clear, measurable understanding of the value their solutions deliver — and which can align their pricing accordingly — will pull away from the competitors who have failed to do the same.
The rise of the 'build, buy or borrow' debate
AI-native systems are not simply incremental improvements on existing technology. They have the potential to fundamentally redesign workflows, automate decision-making and eliminate layers of operational complexity that legacy platforms were never designed to address.
This raises the stakes for carriers. Should they build AI capabilities internally because these are core to their competitive advantage? Or should they buy from a vendor because speed and specialization matter more?
Increasingly, carriers have a third choice: borrow external operating capability through a forward-deployed model. The "borrow" option is driven by the emergence of a new breed of insurtech solution provider. AI-native carriers have started commercializing the automation, data infrastructure and AI-enabled workflows they built for themselves. In this third option, carriers and MGAs "borrow" an embedded operating capability or a technical team from a vendor to help design and deploy AI in their underwriting, claims and operations workflows.
The next chapter for insurtechs will not be defined by who adopts AI first. It will be defined by who uses AI to create a sustainable competitive advantage. The industry is moving from experimentation to economic discipline, and from AI as a feature to AI as a driver of measurable business outcomes.
