The insurance industry is bringing AI into its operations, but the challenge now is scaling it

Artificial intelligence is no longer an experiment for insurers: it is becoming part of day-to-day operations. From underwriting and risk assessment to claims management and customer service, companies are adopting tools capable of processing information and automating tasks that, for years, relied on manual processes.

It is a change taking place in a particular industry: one that handles enormous volumes of data, but also carries the burden of strict regulation, compliance requirements, and processes measured in years, not months. A Deloitte survey of 200 insurance executives in the United States found that 76% had already implemented generative AI in at least one business function. The problem, the same study says, is something else: moving from proof of concept to a real, scaled implementation remains an obstacle for much of the industry.

Economic pressure is adding to the challenge. Swiss Re Institute projects that global premiums will grow by just 1.3% in real terms in 2026, well below the 3.9% recorded in 2025, in a market that is slowing and where non-life insurance is facing increasing pressure.

Against this backdrop, venture capital is also becoming more selective. According to Crunchbase data at the end of 2025, insurance startups had raised close to $3.9 billion in funding from seed through growth stages during the year, less than a quarter of the amount raised at the 2021 peak. But even with that decline, AI and automation continue to feature in several of the sector's most significant funding rounds, particularly in underwriting, claims processing, and risk assessment.

The question, therefore, is no longer simply what AI can do within an insurance company, but how prepared the operation is to incorporate it.

The industry moves from experimentation to scale

The move toward AI is also changing the type of companies attracting capital within the insurtech ecosystem. Crunchbase found that many of the companies funded recently are focusing on infrastructure and workflows, rather than the direct-to-consumer models that dominated earlier stages of the sector.

The reason is tied to the very nature of the insurance business. Automating an isolated task can create efficiencies, but doing so across processes involving different systems, regulatory requirements, and human decisions requires a different structure.

Deloitte found that although 76% of surveyed executives had already implemented generative AI in at least one function, companies still face obstacles related to resources, responsibilities, return on investment, and technological readiness to move these systems from pilots into production.

AI adoption, then, does not eliminate the complexity of the sector. In some cases, it makes that complexity more visible.

That is where a second discussion is emerging within the insurtech ecosystem: how to build operations that can grow at the same pace as the technology.

GetCovered takes that discussion into compliance

GetCovered operates at the intersection of insurance, risk management, and compliance for property managers in the US rental market.

For Rick Folgmann, the company's COO, the growth of an insurtech cannot be separated from the discipline with which it builds its processes.

"In InsurTech, growth without discipline isn't momentum—it's risk."

The company argues that part of that discipline involves building repeatable and scalable workflows before an organization grows to the point where changing them becomes more costly.

"At Get Covered, we focus on building repeatable, scalable workflows early, while the organization is still flexible enough to change," Folgmann said.

The logic becomes increasingly relevant as a platform adds more customers, states, insurers, and revenue streams. In insurance, each expansion can involve new eligibility conditions, compliance requirements, documentation, and customer service processes.

AI can automate part of that work, but it also introduces another layer of responsibility around how those decisions are made and supervised.

Automation does not eliminate compliance

GetCovered's approach is based on bringing compliance closer to day-to-day operations. Rather than treating it as a separate function that steps in at the end of a process, Folgmann argues that it should be part of the infrastructure on which an insurance company operates.

"Insurance regulation isn't a checklist—it's an environment," he said.

Licenses, registrations, disclosures, audits, and responses to regulators are among the processes that can become fragmented as a company grows rapidly. For a platform seeking to automate them, the challenge is not simply to incorporate AI models, but to determine what information they use, what rules they must follow, and when a person needs to intervene.

The same principle extends to the customer experience. Folgmann argues that it does not depend exclusively on an interface or support team, but on the operational decisions taking place behind every interaction.

"Customer experience is often framed as a product or support issue. In reality, it's the output of dozens of operational decisions."

In this context, the evolution of AI in insurance is beginning to change the nature of the conversation. The market is moving from asking whether these tools can automate specific tasks to discussing how they can be integrated into operations that must account for regulation, data, risk, and service.

Funding in the sector also reflects this shift. Although capital going into insurtech remains well below the highs of 2021, Crunchbase identifies a greater concentration of companies using AI to automate processes and build infrastructure for the insurance business.

For insurers and startups seeking to work with them, the next stage may depend less on proving that AI works and more on building the conditions for it to work consistently.

In an industry whose business is precisely about managing risk, the ability to scale technology without multiplying operational complexity is becoming part of the equation.