Property claims adjusting is losing people faster than carriers can replace them. Nearly a quarter of experienced adjusters are expected to retire by the end of 2027, according to Sedgwick’s 2026 Loss Adjusting Insights Report, and the ones leaving are the ones who know the most. The policy knowledge and judgment that used to bring a new adjuster up to speed is walking out the door with them.
When an adjuster leaves, their caseloads get transferred to those who stay. Heavier caseloads bring more potential mistakes, especially on routine work like interior measurements, which eats an afternoon and should never have depended on a tired adjuster’s tape measure in the first place.
How can AI help with the claims adjuster shortage?
AI can optimize the workflow adjusters already use by speeding up its slowest steps: capturing the property, measuring it and drafting the estimate. The adjuster’s process does not change, and neither does the adjuster’s authority. A person checks the work at every step, and the adjuster makes every coverage decision. AI used this way adds capacity to a claims team without removing the human judgment the job depends on.
Most adjusters don’t trust the AI tools they’ve been handed. Of the Glassdoor adjuster reviews that mentioned AI over the past year, 98% were negative. Part of that is fear of job loss, but much of it is practical application—the tools adjusters have been handed get things wrong, and the adjuster is the one left cleaning up. One described losing 15 to 20 minutes per file correcting AI errors that nobody had reviewed.
At Elara Spatial, we are pro-human about AI, and we built our platform to fit inside the adjuster’s existing workflow, with human oversight required at every stage, rather than to work around the adjuster. Here are the four places in that workflow where AI can save time, and where a person stays in charge of each one.
1. Capturing the property
The workflow today goes something like this: the claim is opened, the adjuster schedules a site visit and the policyholder waits. For a routine water loss, that wait can run weeks.
The optimized version keeps every one of those roles and removes the waiting. The carrier sends the policyholder a link by text or email, and the policyholder photographs each affected room on their own phone, guided step-by-step so the images come out right the first time. No special device and no app are needed.
The adjuster still decides which claims need an in-person inspection. A burst pipe in a finished basement, a kitchen fire contained to one wall, or a roof leak that stained two ceilings can be captured by the policyholder the same day, and the adjuster’s first look at the damage arrives as photos and a 3D model instead of a date on the calendar.
2. Measuring the property
Measurements are the most routine part of a claim and the easiest to get wrong at the end of a long day. They are also the step most adjusters would gladly hand off, provided the numbers can be trusted.
AI can rebuild a room in 3D from ordinary smartphone photos and pull the dimensions, materials and damage from that model. The measurement step in the adjuster’s workflow stays where it is. It simply arrives completed, and the adjuster reviews it rather than produces it. If something looks off, the photos and the model are there to check against.
3. Drafting the estimate
Drafting an estimate line by line, to the carrier’s own guidelines, is where most of an adjuster’s hours go. It’s also the step that most needs a human signature.
Our platform drafts the estimate to each carrier’s guidelines, and then a human expert reviews every line before it goes anywhere. Only then is it delivered in Xactimate, where the adjuster reviews it again and adjusts it as they would any draft. Across our deployments to date, the time from loss to a damage repair estimate has dropped from about 28 days to about two hours, and more than 80% of estimates have arrived error-free. The adjuster’s judgment is still the last word, and it now goes toward reviewing an estimate rather than building one from scratch.
4. Showing its work
Most AI tools fail adjusters here. An estimate the adjuster can’t trace is an estimate the adjuster has to redo, and that’s where the 15-to-20 minutes a file comes from.
Oversight only works if there is something to oversee. We built our platform as an assembly line of narrow AI models, each doing one specific job (identifying a material, for example), with ordinary software passing the work from one step to the next. Every step is logged on every claim. If a line item is questioned, by an adjuster, a policyholder or an attorney, the record shows how it was decided. That is what makes human review meaningful rather than ceremonial, and it is what lets an adjuster defend an estimate they did not draft from scratch.
Where the adjuster stays in charge
The hardest part of an adjuster’s job is often the conversation with the policyholder. For many homeowners, the adjuster is the first person who walks them through what happens after a loss, and that soft skill is different from reading a floor plan or pricing a repair—but both are expected of the same person. When an adjuster’s day is spent measuring and drafting, there is less of them left for the policyholder.
We believe that part of the job must stay human. A frustrated homeowner or a policyholder’s attorney needs an adjuster who can explain a coverage decision and stand behind it, and no software we know of can do that well. What AI can do is give back the hours that measuring and drafting take, so there is more of the adjuster left for the conversation. With a person checking the work at every step, a smaller team can handle the caseload of a larger one without being asked to trust a number nobody looked at.
Frequently asked questions
Will AI replace claims adjusters?
Some AI tools may be built to try, but Elara Spatial is not one of them. Our platform is designed to work inside an adjuster’s workflow, never in place of it: AI handles the capture, measurement and drafting steps, while coverage decisions, negotiations and policyholder conversations stay with the adjuster. Every AI-assisted step is reviewed by a person before it moves forward, and that is a design choice, not a setting.
Does using AI change how an adjuster works?
It can, and that’s usually the problem. Many AI tools ask adjusters to learn a new system or work around the tool’s output—but that’s not the case with Elara Spatial. Elara fits inside the workflow adjusters already use: the estimate arrives in Xactimate, where the adjuster reviews and adjusts it as they would any draft.
The adjuster’s workflow keeps the same steps in the same order. Elara shortens the three that used to take the longest: capturing the property, measuring it and drafting the estimate.
How is an AI-generated estimate checked?
Twice. A human expert at Elara reviews every estimate before it is delivered, and the adjuster reviews it again in Xactimate. Every step of the process is logged, so any line item can be traced back to how it was decided.
Does the policyholder need to download an app?
No. The carrier sends a link by text or email, and the policyholder opens it on any smartphone and is guided through photographing each affected room.
Elara Spatial fits inside the workflow adjusters already use, turning a few smartphone photos into a line-by-line damage repair estimate that is reviewed by a human expert and delivered in Xactimate for the adjuster’s final review. Learn how it works.