The Future of Claims: Preparing for Claims Management 2.0

Claims management 2.0 was a forecast before it was a description. In the spring of 2023 the industry had settled on roughly five bets about where claims work was heading, repeated in enough conference sessions and vendor decks to count as a consensus: digital would become the default, AI would arrive, communication would get more transparent, customer centricity would stop churn, and the operations that kept up would be the ones still standing.

Enough time has passed to mark them. Several arrived exactly as described. One arrived and immediately changed the question underneath it. One did not arrive at all, and its absence is the most useful item on the list, because it points directly at where a claims operation should be spending effort now.

The five bets, marked

What follows is each one, and why the verdict matters more than the prediction did.

Digital arrived, and turned out to be the floor

The 2023 case for digital was demographic. PwC put Generation Z and millennials at around 38% of the workforce at the time, heading for 58% by 2030, and the argument was that their expectations would drag insurance along behind them.

The demographics held. The strategic conclusion drawn from them did not survive contact with the market, for an unremarkable reason: everybody acted on it at the same time. Digital first notice, a claimant portal, document upload from a phone, and electronic payment are now assumed rather than compared. A capability every competitor has is not a strategy. It is a prerequisite, and prerequisites are worth exactly nothing in a competitive conversation.

So the interesting question moved one step downstream. Not whether a policyholder can file a claim on a phone, but whether anything visible happens after they do. That is bet three, and bet three did not land.

The AI forecast became a description, and the hard part moved

The 2023 evidence for AI was a survey. An NAIC study found that 89% of auto insurers either used AI and machine learning, planned to, or were exploring it.

Read the wording again, because it is the whole story. Used, planned to, or exploring. The number counted intent. Intent is a reasonable basis for a forecast and a poor basis for a purchase, and the gap between the two is where most of the last few years of AI in claims actually happened. Adoption is no longer the line that separates claims organizations. Nearly everyone has bought something. The line now runs between operations where AI touches work people already do and operations where it sits beside the work, impressive in a demonstration and quietly unused on a Tuesday afternoon.

The other half of the 2023 argument was cost. It usually arrived as a consultancy estimate of how much of a claims journey automation could remove, and cost is the easiest benefit to model and the easiest to lose again. An automated step that produces a wrong reserve, or a coverage note a client quotes back months later, costs more than the handling time it saved. In claims a wrong answer does not stay wrong quietly.

VCA's answer to that was to wait, and to say so out loud while the market was at its loudest. The company watched how claims organizations were actually using these tools, asked its clients what they wanted from them, and then started where the work already existed: the documents adjusters upload anyway, and first notice, because a claim opened properly sets the tone for everything after it. The rule underneath is effectiveness before efficiency. Automation is only good when the result is right. AI finds, summarizes, and drafts; the claims professional decides, sets reserves, approves, and authorizes payment. See how that line is drawn, and where it stops.

Transparency was the cheapest bet and the least finished

This is the one that did not arrive, and it was never a technology problem.

In its 2025 study of the digital claims experience, J.D. Power found that insurers deliver adequate proactive updates through digital channels only 22% of the time, even though those updates rank among the strongest drivers of claims satisfaction.

Sending a status update is not hard. Knowing the status is hard. A claimant portal can only display what the operation can see, and what a claims operation can see is routinely less than what it answers for.

That is the shape of the problem for anyone running claims across a panel, a program, or a set of carrier clients: you are accountable for claims you cannot always fully see, handled by people and partners you do not control. Every handoff is a place a claim can stall, and most handoffs happen outside the system that is supposed to be managing it. Until the handoff is in the file, transparency is a promise the front end cannot keep. The digital self-service side of this has its own set of numbers, and they all point the same way.

Customer centricity turned out to be an operations problem

In 2023 this was framed as a philosophy: make the customer the focal point of every decision about products, services, and experiences. Stated that way it is unarguable and unbuildable. Nobody has ever gone into a quarterly review and reported how much focal point they delivered.

What actually moves a claimant's experience is duller and much more specific. Whether the next step on their file lives in the file or in somebody's inbox. Whether the adjuster who picks the claim up on Tuesday can see what was promised on Friday. Whether the person who calls to ask for an update has to wait for someone else to call a third party first. Those are workflow questions, and they are answerable.

The misdiagnosis had a cost. Organizations that treated churn as a culture problem ran satisfaction training. Organizations that treated it as an operations problem fixed the assignment, the next step, and the record of what changed. Only one of those two shows up in a claims file.

The bet that aged worst was "keep up"

The fifth bet said the risk was falling behind. Move fast, adopt, or face the churn.

That framing put every claims leader in the position of proving they were not late, which is a poor state of mind for buying software. The risk that actually materialized was the opposite one: adopting faster than the operation could govern. Features configured because they were included. Modules switched on because a competitor mentioned them in a webinar. Automations nobody could explain to an auditor eighteen months later, in a business where explaining yourself to an auditor is the job.

The healthier posture is restraint as a design principle rather than caution as a personality trait: configure what matches the work, skip the rest, add later. Skipping is the part that gets left out of the pitch, and it is the part that keeps a configuration legible to the next person who inherits it.

What none of the five bets accounted for

Look at the five together and one assumption runs through all of them. Each bet describes something an insurer does to its own process, inside its own walls, on its own schedule. Digitize your intake. Adopt your AI. Improve your communication. Center your customer. Keep up.

The work does not sit there. Claims do not happen inside systems, they happen across organizations: across teams, partners, programs, and reporting obligations. A claim moves from an insured to a carrier to a TPA to an independent adjuster to a vendor to a bordereau, and the places where it stalls are almost never inside one of those organizations. They are in the space between two of them. That is why claims software is better judged by how well it moves a claim forward than by how well it stores one.

None of the five bets were wrong about their own subject. They were all scoped to a single organization in a business that is not run by single organizations, which is why doing all five well still leaves a client asking who has this and when will I hear again.

Where the effort goes now

Four things follow from the marking, in the order they are worth doing.

1. Fix what you can see before you buy what predicts. Visibility across the parties on a claim is the cheapest unfinished work in most operations, and it is the input every other improvement depends on, including the AI ones.
2. Judge an AI capability by its review step. Ask what a person does with the output, where it is attributed, and what shows up in the activity log. A capability with no answer to that is a demonstration, not a tool.
3. Ask any vendor what is shipped and what is roadmap, in writing, with dates. You will learn a great deal from who hesitates. The 2023 forecasts were graded eventually; product claims should be too.
4. Count the cost of adopting, not just the cost of licensing. Time to live and time to competence are the numbers that decide whether anything on this list actually happens.

On that last point, VCA publishes its own bands rather than leaving them to the sales conversation: Standard License deployments go live in 3 to 4 weeks, Enterprise License deployments run 8 to 16 weeks, Carrier License deployments run 12 weeks to 6 months, and file handlers typically need 30 to 60 minutes of training.

Frequently asked questions

What did the claims industry predict for claims management 2.0, and how did it turn out?

In 2023 the industry bet that digital would become the default claims channel, AI would take off, communication would become more transparent, customer centricity would stem churn, and keeping up would be the survival skill. Digital arrived and stopped being a differentiator, and AI arrived but the hard part moved. Transparency did not arrive, customer centricity half arrived and was misdiagnosed, and the keep-up bet aged worst of the five.

Why is digital claims intake no longer a differentiator?

Digital first notice, a claimant portal, document upload from a phone, and electronic payment are now assumed rather than compared, because everybody acted on the same case for digital at the same time. A capability every competitor has is a prerequisite, not a strategy. The question has moved downstream, from whether a policyholder can file a claim on a phone to whether anything visible happens after they do.

Why do claimants still not get proactive claim updates?

In its 2025 study of the digital claims experience, J.D. Power found that insurers deliver adequate proactive updates through digital channels only 22% of the time, even though those updates rank among the strongest drivers of claims satisfaction. Sending a status update is not hard; knowing the status is. A claimant portal can only display what the operation can see, and most handoffs on a claim happen outside the system that is supposed to be managing it.

Claims management 2.0, stated plainly

The 2023 version of this argument was a list of things to get ready for. The version worth holding now is shorter and less exciting.

Digital is the floor, not the differentiator. AI is worth having where it works on documents and intake your team already handles, and worth refusing where it decides. Transparency is an operations capability wearing a customer-experience costume. Customer centricity is assignment, next step, and record. And keeping up is not the skill. Knowing which of the five to leave alone this year is the skill.

The claims operations that look modern in 2026 are not the ones that adopted the most. They are the ones that can answer, on any open claim, who owns this, what happens next, what is waiting and on whom, what changed, and when the client will hear again. Everything else on the 2023 list was in service of those five questions. It just took a few years of building to notice.

If you want to see how those questions get answered on a claim like yours, ask for a walkthrough.