AI is changing workers’ comp and small commercial – but how will insurers build judgement?

For workers’ compensation claims professionals and small-commercial underwriters, some of the most important changes brought by artificial intelligence may not be about replacing people at all. They are about changing what people spend their time doing – and, in turn, how they learn the insurance business.

That was a central theme of an Intelligent Insurer  webinar held in partnership with Markel, where Reece Adams and Jaime Thompson of Markel, Irina Simpson of Gallagher Bassett and Jeffery Kaczynski of Sollers Consulting examined how technology is reshaping workers’ compensation and small-commercial insurance.

AI can summarise information, identify patterns and automate routine tasks across underwriting and claims. That can free professionals to focus on more complex risks, customer relationships and decisions requiring experience and judgement.

But it also raises a difficult question for an industry that has traditionally developed expertise through repetition: if technology handles more of the routine work, how do the next generation of underwriters and claims professionals gain the experience needed to make good decisions?

Moving underwriting up the value chain

For Adams, managing director of Underwriting, Workers Compensation and Small Commercial at Markel, AI is already changing the role of the small-commercial underwriter.

Rather than spending as much time gathering data and working through high volumes of relatively routine accounts, underwriters can increasingly focus on interpreting information, assessing risk and determining the right strategy for an account.

“What has changed is that underwriters are moving up the value chain,” she said. “The role is moving from data gathering and heavy, volume-based process work to much more significant analysis of risk, account strategy, cross-selling and, ultimately, more meaningful agency conversations.”

That shift has particular significance in small commercial, where insurers deal with large volumes of accounts and need to balance speed and efficiency with the expertise required to understand individual risks.

 “Underwriters used to gain judgement by looking at thousands of accounts. That is how you learned.”

The technology can handle more of the processing, but the underwriter still needs to understand why a risk should be accepted, declined or approached differently – and be able to explain that decision to an agent.

That is where the industry's traditional approach to developing judgement becomes more difficult.

“Underwriters used to gain judgement by looking at thousands of accounts. That is how you learned,” Adams said.

If AI is increasingly performing the first layer of analysis, junior underwriters may have fewer opportunities to build that experience simply by working through account after account.

Markel is therefore exploring ways to make the learning process more deliberate. Less experienced underwriters can analyse a risk before seeing an AI-generated summary, for example, and then revisit previous declinations to test whether their own assessment changes when they see the technology's analysis.

“The training model certainly looks different, but the crux of it is coaching how to think, and how to challenge a model,” Adams said.

That ability to challenge an AI output could become as important as knowing how to use it.

Human relationship remains critical

The implications are different but equally significant in workers’ compensation claims.

Simpson, executive vice president of workers’ compensation claims at Gallagher Bassett, said technology was forcing the industry to reconsider how it develops talent at a time when insurers and claims organisations are also facing significant workforce pressures.

“This technological revolution is going to force us to reassess what we are doing for the future,” she said.

AI can help claims professionals process information more efficiently and identify important details within large volumes of claims documentation. But workers’ compensation also involves something technology cannot easily replicate: the human relationship with an injured employee.

A worker may be dealing with an injury, uncertainty about their recovery and concerns about their ability to support their family. In those situations, Simpson said, efficiency cannot be the only measure of success.

“There is no chatbot that is going to alleviate that,” she said. “That is where you want a human being who is available, who is not hurried because they have another 130 cases to handle, who can dedicate that time to you.”

The opportunity for AI is therefore not simply to reduce the number of people involved in claims. It can allow claims professionals to spend less time on administrative work and more time on the conversations and decisions where their expertise matters most.

Using AI where it adds value

Kaczynski, manager and US Workers' Compensation Leader at Sollers Consulting, pointed to practical examples of how AI and connected workflows can support workers’ compensation professionals.

Technology can summarise information sent to underwriters and identify medical conditions buried within claims notes that could affect an injured employee's return-to-work plan.

But he cautioned insurers against assuming that increasingly sophisticated AI is always the answer.

“AI is not the silver bullet. It is a tool,” Kaczynski said. “You pick a proof of concept and you use it. It is an effective tool, yes, but it is not the only thing in your box of goodies. At the end of the day it is part of the human experience.”

For workers’ comp in particular, the starting point should be the business problem. A relatively simple automation may sometimes deliver more value than an AI agent, he said.

That means insurers need to develop not only technical capability but also AI literacy among the people using the technology.

“We have to improve AI literacy across the workplace,” Kaczynski said. “A lot of people are wary of using AI, so let them experiment and run classes. Then upskill your underwriters and claims specialists to use it effectively.”

Replacing volume with experience

The question of how workers’ compensation professionals build expertise is already prompting new approaches to training.

Gallagher Bassett's six-month Invest graduate programme selects 30 candidates and focuses on relationship building, decision-making, problem-solving and technology fluency – capabilities Simpson said reflect how the role is changing.

The programme also seeks to recreate some of the case experience that traditionally came from working through large volumes of claims.

“Where you previously had to go through a thousand files, here is a set of scenarios you can be taken through in a customised pathway,” Simpson said.

The approach illustrates a broader shift taking place across workers’ comp and small commercial.

If technology handles more of the routine cases and administrative work, insurers cannot assume that expertise will simply develop organically through exposure to volume. They may need to create more deliberate opportunities for employees to assess scenarios, challenge assumptions, make decisions and learn from the outcomes.

For small-commercial underwriting, that could mean giving junior underwriters structured opportunities to form a view of a risk before seeing an AI assessment.

For workers’ compensation, it could mean using simulated claims scenarios to expose new professionals to the range of decisions and human situations they will encounter.

In both cases, the objective is the same: use technology to accelerate learning rather than allowing it to remove the experience through which learning traditionally happened.

The next generation of insurance leaders

That puts greater responsibility on managers, Thompson, managing director of Business Development, Workers Compensation and Small Commercial at Markel, said.

As technology supports more of the preparation and analysis, leaders will need to spend more time developing employees' ability to make decisions and articulate their reasoning.

“Leaders are going to have to step up and take on the role of coach and mentor,” she said. “We will have to think about professional development very differently. We will let technology support the preparation but still require the individual to form a point of view and explain why they reached the conclusion they did.”

For small-commercial insurers, that could mean a shift away from measuring an underwriter's contribution simply by the number of accounts processed.

For workers’ compensation, it could mean measuring success not just through claims efficiency but through better outcomes for injured workers.

In both areas, technology creates an opportunity to move people towards higher-value work – but only if insurers are prepared to rethink how roles, training and performance are designed.

“I would stop measuring success purely through efficiency and continue investing in judgement, expertise and relationships,” Thompson said. “It is critical that we get the customer and business outcomes right first.”

For workers’ compensation and small commercial, the AI challenge is therefore not simply about automation. It is about making sure that the people who remain at the centre of underwriting and claims have the skills and experience to use that technology intelligently.

The insurers that get this right will not necessarily be those that automate the most. They will be those that use AI to remove routine work while deliberately creating new ways for underwriters and claims professionals to build the judgement, expertise and relationships that remain at the heart of insurance.

Did you get value from this story?  Sign up to our free daily newsletters and get stories like this sent straight to your inbox.