You Know Who Hates AI? Insurance Claims Adjusters
wired.com
wired.com
As an analogy, I don’t know if I fully blame my executives for buying Microsoft Teams when Microsoft Teams turns out to suck. Microsoft promised to my company’s executives that Microsoft Teams doesn’t suck just like Sam Altman promised that AI is so powerful and so capable of exceeding the capabilities of humans that it’s dangerous.
Maybe at some point some of the B2B customers of AI products will be suing their vendors for the kind of grossly exaggerated marketing we are seeing right now. I think that SaaS companies would never get away with it.
I don’t know that the executives at my company suddenly lost their mind and decided that hundreds of dollars of AI spend per employee is suddenly justified even though approving a simple $##/month/employee SaaS tool was previously a rigorous slog. The only reason the executives at my company suddenly changed their mind on frugality is becausey they were sold a lie by a huge industry.
We could consider them clueless for falling for that, or an alternate viewpoint is that they are victims of widespread fraud.
Nobody is forcing you to take poison, and saying "Well Microsoft promised me it isn't poison!!!" at this point, decades into it selling poison, doesn't make much sense.
You are not correct.
Literally anyone off the street can be a "decision maker" in technology today. Just buy what everyone else is buying and collect a huge paycheck while working remote. People who understand tech and sort of know what they are doing, are not included.
That has been my experience.
When he asked me to email him scans of a bunch of personal documents, I asked him for his PGP key.
..."What's that?"
Of course, if insurance companies were rapidly held to account by courts, a core function of the courts if there ever was one, there would be no problem here.
Which is probably why most governments are defunding the courts.
Somehow the "accountability" shifted from meaning "being reaponsible for not defrauding, lying" into "it is fault of people who were lied to".
It's like a person voting for Trump the second time being confused at all his lies, at this point basic education would get you out of this.
Pretty much what this blog post says: https://hermit-tech.com/blog/ai-mania-is-eviscerating-global...
It seems pretty bad, no doubt about it
This is what they depend on actually. Scape goats for their failures. This quarter takes a hit then number go up.
If so, yes, they are to blame, either for not being able to evaluate the stuff they buy or for deciding on questions they are not qualified for. Either one is incompetence.
You can also blame them to continue renting the thing (because it's not brought) after it's shown to be defective. And if they decide to do irreversible transformation to your business before verifying that the thing even works, than yes, that's bare incapacity of running a company.
Half my colleagues are besotted with the technology and unaware that it is shaping their decision making, not the other way around.
Hard to fathom a modern model would get any of this basic stuff wrong. I know accountants using it for (a lot more complex but similar problem) tax law parsing and it going through flawlessly.
There seems to be some bidirectional media gaslighting happening here where the doomers and glazers are talking past each other a bit.
You can't 'blame' the AI, as it can't be held accountable.
That makes it way more likely that people blame AI (doesn't make it right by any means, but does mean that the actual blame gets diffused even more).
1) employee was willingly sabotaging the company: employee gets jail time plus fine, employer is on the hook financially
2) employee was involved either in accident or just made a stupid decision: employer is on the hook financially
So if that's the case, then employers are financially responsible for everything the AI they use says. That's the rule for employees.
Note: this is different from, say if a company uses a car and it just starts by itself and drives into the next door warehouse setting it on fire (ie. some sort of designed-in decision/issue) : in that case the maker of the car is responsible, both financially and criminally, even if an employee or even the boss/owner/everyone at the company were involved.
Of course, guaranteed that governments will change this the first time a court makes this obvious connection. Why do I think that?
Well, for example, the UK government changed the rule when it turned out the post-office's written software (written by contractors, under post office and thus government responsibility and accountability) decided to cause incredible damage by blaming innocent people for stealing. Then, a special law was voted that in that specific case it didn't apply, and the government department, nor any of it's employees, were responsible.
Just like, for example, Pennsylvania changed the rules for unwittingly-but-directly aiding criminal activity when it turned out half the Pennsylvania government was complicit in the kids-for-cash scandal (because in that case, normally, you wouldn't be criminally liable, but you WOULD be liable for any financial damage you do)
In order to learn and improve you have to know what went wrong, with AI being personified and pulled into the pool of entities that may be at fault, it makes it harder to figure out what actually went wrong and learn. Not impossible to figure out, just harder.
Seems like more of a semantic distinction IMO. Yes, I can't technically "blame" the AI because it's not accountable. But what word would you suggest for "I had [X technology] handle [Y task] and it failed to perform that task?" I can't "blame" my router if I lose internet and it prevents me from jumping on a Zoom call, but it's also true that the router failed to do the thing it was supposed to do.
When my internet goes down, I blame either my ISP for having an outage, or the manufacturer of my router (assuming it's a router issue? Haven't had that happen personally but I'm certain it exists) (Or i did something dumb with my ufw again but that's on me)
If you create an autonomous system and it fails, blame should be on you, imo.
The word is still 'blame', it just needs to be applied correctly.
LLMs aren’t there yet, but they might get there. My point is that your descriptions fail to capture when or why exactly the blame would shift.
So they Google "Insurance claim AI tool", land on a vibecoded SaaS that is just a wrapper on a pocket Chinese model spun as "your next insurance pro", and then are getting the whole department on some lone 19 yr olds weekend project.
Another part of the problem is that a model not specifically fine-tuned to make a total loss determination won't know the relevant factors, nor how an insurance company's concept of a total loss differs from the public's.
And still another part of the problem is that most total loss claims aren't what you, dear reader, are imagining: They are very rarely "the car is a thin pancake after being crushed by a meteor".
The much, much more common scenario is: "50% of the body panels sustained at least paint damage, both headlight modules need replacement, and the front wheels look funny. Given that the vehicle has an MSRP of $FOO, $BAR miles, no prior collision history on carfax, and is a popular color, is it cheaper to repair or total the vehicle?"
Of course, the model can turn over the hard cases to a human adjuster... but then what are we doing here? It only takes 10 seconds for the human adjuster to handle the "crushed by a meteor" case also.
Source: Listening to my SIL rant about being asked to stop training bespoke total loss models and just send it by 1-shotting a commercial LLM.
Not saying they're not good for other things, but the "waste people's time and make them go away" use is just so perfect for the capabilities of even cheap models.
You see, I live with in a rural area and especially since LED light bulbs have become the thing, it's been clear that we are delivered far less than 120v service a lot of the time. We live about 200 yards from the transformer that we share with several other houses, too many I believe. About two years ago I set up Home Assistant and bought one of those power monitoring rigs that you put on each circuit in your breaker box, mostly I just wanted data about where my power bill money was going. I work in FinOps in case anyone was curious.
Anyhoo, couple years later I'm analyzing several months worth of power bills vs the circuit meter readings to quantify where I can make some upgrades and save some money. After a day or so of this I mention to Claude about the voltage issues and sure enough, it puts together a whole packet for me documenting brownouts down to 95v and routine brownouts below 110v. It helps me fill out the BPU complaint form.
2 days later I get a call from JCP&L (unfortunately to my wife's phone because it's her name as the primary on the account). I want a new, larger transformer for myself and my neighbors, so I need to call them back today.
I’ve seen half a dozen houses on a single 7.2kV -> 240V 25kva transformer which can lead to problems given the available current on the secondary side is only 104A. As the transformer overheats you get voltage drop, as you’ve noticed ;)
Your supply voltage should be +/- 5% of nominal, 115-125V line to neutral and 230-250V line to line, appliances are generally OK with +/- 10% from nominal voltage.
Like I said, it routinely drops to or below 100v. It's most noticeable after an outage is restored when everyone's hot water heaters and fridges kick back on at once.
The article is entirely about the AI doing the first pass of the claim intake and causing a lot more work for everyone involved instead. Mis-routing, mis-categorization, hallucinating details, etc.
It used to be you just email them a photo or video of the issue and they send you a replacement. Now companies won’t send one until they have received the defective one back.
Then again my last two online purchases, a pair of shoes arrived stained with somebody else's socks in the box and a cordless vacuum arrived with a mismatched brush head for a different model.
It's interesting how companies are now redefining things they explicitly encouraged their customers to do as "abuse."
Some also gets diverted into return resale stores where companies like Amazon will sell a 3rd party a pallet of miscellaneous returns they then sell.
Even then, the one clothes example (shoes - which had been returned stained and with socks in the box) mentioned in the above still wouldn't seem like a valid return by any stretch.
They sell clothes and also shoes.
Zalando very much is operating on the principle that you order things to try them out at home and return whatever doesn’t fit or whatever you end up not liking. Just like in a physical store you try out clothes and shoes inside the store to see how it fits and how it looks on you.
That being said, there are limits to what they consider reasonable:
> One key part of that experience, and something our customers deeply value, is the ability to make an order, try the items at home, and, if needed, easily return those they don’t want to keep. In short: bringing the fitting room to your home.
> […]
> Customers whose return activity reaches disproportionately high levels will receive an email warning. This proactive step gives them the opportunity to reflect on their behavior, discover the solutions we provide to avoid returns, and adjust before further measures are taken.
https://corporate.zalando.com/en/convenience/bringing-transp...
As someone with wide feet, I don’t buy shoes unless they specifically have width options. It’s not particularly easy to do unless you really like traditional men’s leather shoes and boots, but there are some athletic shoe makers that offer wide width shoes (NB and Hoka for sure).
Even given all of the above, I still own shoes in 3-4 different sizes because shoe sizes are not standardized between makers, like you said.
Dealing with returns is the tradeoff that online sellers make in order not to put a store in every city or to deal with retail arrangements.
Not so for consumer electronics
Oh yes it is. Buying something you know you don't plan to keep is abuse. If you ran a store, would you be fine with 2/3 of your orders being returned? A good litmus test for "am I being an asshole" is "what would the world look like if everyone behaved like me?"
Now whether it's a good thing wrt the environment etc...
Buying 3 new-in-box $1000 cameras and returning 2 of them covered with your smudgy fingerprints is abuse, no matter how you slice it. If everyone did it, those cameras would cost twice as much. I'm saddened but not surprised people are defending that behavior here.
I agree. What percentage of online returns do you think is similar to this scenario?
What percentage of "buy N and return N-1" return patterns (which you believe to be abusive) is people buying multiple $1,000+ pieces of electronics equipment, versus say 3 sizes of shoes or 3 colors of a dress?
My guess is the former case, which I agree is abusive, is <1% of such return patterns. Partially because it's a legitimately stupid way to shop for those types of items and requires significant credit, while buying multiple sizes/colors of items like clothing is a completely sensible way to shop.
Most online shops ToS say "we can ban you at any time for any reason", so their lawyers' definition of the word wouldn't be relevant here anyway.
That kind of argument only works where stopping the behavior is cost free, the main reason most places didn't do it before was because the number of people doing it was low so the cost of actually tracking returns rigorously outweighed the benefit of low friction returns for the 99+% of customer who weren't abusing the system. When more people start creating large numbers of returns suddenly the amount of losses they create become enough to care about stopping. (Or it becomes a low hanging fruit to squeeze more profit out of an existing business, spending customer preference/loyality for profit by cutting a small waste center)
When your business model has gotten to the point where people cheer for a guy who murdered one of their CEOs, it’s not surprising that people would not be interested in joining and may see being murdered as an occupational risk beyond what their personal actuarial tables allow.
On the outside, I was able to use a consensus model to reach a successful settlement. I wouldn’t say I won, but was made whole after having to file in court.
If AI is making their lives miserable on the inside as well, I fully support it.
And all the while these crooks are claiming it's so humans can focus on more important things. Yeah right. If we're lucky, AI won't improve and there will be a huge backlash, like outsourcing in the early 2000s. If AI gets better, actually good enough to do most work currently done by humans, well then probably something really ugly happens.
Yes it's ALSO to a promote a book, but still... there's a lot more to it than just mindlessly raging against a new technology.
[1] https://99percentinvisible.org/episode/552-blood-in-the-mach...
People have no idea how inefficient the U.S.’s healthcare system is. And it isn’t just the private companies. A ton of states purposefully overburden and underfund their Medicaid systems for ideological reasons. Then the private insurance companies adopt the same practices later.
> These reports are frequently over 100 pages and she is only slotted so much time to do it, so she ends up doing a ton of unpaid work.
> A ton of states purposefully overburden and underfund their Medicaid systems for ideological reasons.
Then her client doesn’t get the help they need? That should be obvious.
> This seems like an open-and-shut case with the US Dept of Labour.
Even if it wasn’t for the “u” in labor I could tell you don’t live in the US.
Otherwise the situation is more complex than the original comment suggests, and I can only speak to one state (Idaho), but I don't think the scenario is uncommon in other GOP controlled states and isn't limited to Medicaid.
Idaho has cut taxes in the state for five consecutive years, largely (IMO) on an ideological basis. This had the result, desired by many in a deeply conservative legislature, of creating budget deficits. Even the GOP is normally wary of espousing cuts directly to the state's contribution to Medicaid, but the deficit gives them cover to reduce funding and introduce barriers to service like caps on authorized services and new work requirements. Even when revenues were higher than expected for FY 2026, the governor's office had to act to actually retain those unexpected revenues as it was pretty clear that the Legislature was going to look to claw them back and continue drive down State spending. It's worth noting that health services, including Medicaid funding, represent about a third of the State's budget and their largest portion of the budget.
This whole process has been playing out in public education funding in the State for years as well. Coincidentally, charter school enrollment has increased by roughly 35% over the same 2021 - 2026 period.
Attacking a policy, program or institution financially and then pointing to the deficiencies of those same programs to further fuel funding and efficacy reductions is a pretty well established tactic in US politics as I'm sure it is elsewhere.
Depending on the competition for her job and the state laws, she would either be put onto a performance management program, or fired. I've been through this before. The HR approved response is that they expect those 100-page reports to be complete within normal working hours, and a failure to achieve that means you are unable to fulfill the requirements of the job.
Of course those in high demand who can easily move to better employers will often do so. However there is a lot of friction when moving jobs. It's very time consuming, mentally demanding, and risky. I've seen incredibly talented staff treated awfully in my years, and they refuse to leave for various reasons.
EDIT
This part should have appeared first: I agree that your scenario is the most likely outcome, even if I think it is reprehensible!
Except shit, amoral behavior, backstabbing, lying at all times, just to get those sweet money and power.
Now I am in the claims process with the large insurance co., using Claude to basically run my side of the claim — reading the policy, organizing and filing the ALE receipt scans, reading and notifying me of messages from the agents, calling out when they are delaying too much or not fulfilling their promises, challenging their decisions on what to cover based on the policy, keeping a log of all of our correspondences, logging work estimates and invoices, I cant imagine doing all this tedious work myself—w/o Claude I’d probably just rollover fatigued.
Does my adjuster like this? On the one hand, I’m giving her incredibly tidy and documented receipts and logs, which I am sure is saving her time.
On the other hand, Claude is calling out her BS and delays, which have been atrocious—so much so that Claude is now adamant that she has gone too far and that I should be filing a complaint with the state insurance bureau.
> On the other hand, Claude is calling out her BS and delays, which have been atrocious—so much so that Claude is now adamant that she has gone too far and that I should be filing a complaint with the state insurance bureau.
This is an interesting customer strategy. Have you escalated your concerns to her management? Have you explained to her that you are considering filing a complaint with the state insurance bureau? It's the ol' stick vs carrot. I say: Start with the carrot. If that doesn't work, back to the stick!
Yup - by checking your Claude usage meter.
Secondly, does anyone actually know how much of your insurance premium is going on the call centre staff? most people don't claim, those that do mostly have simple claims. The actual cost for insurance companies is actually paying out claims, so even if you automated all of this it likely doesn't change your economics. Historically "New" Insurance companies only really succeed by mis-pricing risk and gaining market share that way - they often then fail when that risk materializes. Lemonade sounds a lot like that.
Finally, they haven't engaged with the tidal wave that's coming. Sure, AI agents handling claims is happening now. Just you wait. In a couple of years time it'll be AI agents making the claims for you and suddenly there'll be bots filing claims with infinite patience and a direct mandate to try and get as big a payout as possible. That is when it's going to get really hairy.
It’s pretty easy to figure out. Insurance companies publish a loss ratio. That’s the ratio of premiums collected to claims paid out.
Lemonade has industry-leading loss ratios now BTW.
The flood of nonsensical AI slop applications that HR is facing has forced this.
This goes for HR, education, code review, on and on.
Yeah sure. I mean, same reason why HN readers hate AI, right? It replaces their jobs.
(Obviously the difference is that the poor old claims adjusters didn't spend the last two decades "disrupting" every other industry and telling everyone else to "learn to code". Ha! If you hear a faint high noise, that's the sound of the rest of the world playing a violin, made of highly disruptive nanomaterials.)
https://www.hawley.senate.gov/hawley-chairs-hearing-that-exp...
The California Department of Insurance characterized State Farm's tactics as "burying survivors in red tape at the worst moments of their lives"
https://www.insurance.ca.gov/0400-news/0100-press-releases/2...
Indeed, some of the core violations could be simply mandated as part of the inviolable system guardrails of an LLM - e.g. failed to begin investigating claims within 15 days, failed to accept or deny claims within 40 days, and failed to pay accepted claims or provide written notice of the need for additional time within 30 days, as required by law.
And at the end of the day, whatever weighting issues an LLM for Insurance Claim Adjustment has, it certainly won't steal the money and run off to a Casino with it
https://www.nicb.org/news/regional-news/broward-insurance-ad...
Anecdotally, many of them are aging out and the demands of adjusters are increasing while the pay hasn’t necessarily gone up. While technically you don’t have to know the ins and outs of construction, it helps a lot and there are just not a lot of people who have that knowledge. And if they do, they often go into construction themselves or sell roofs. Both of which can be far more lucrative.
On top of that, claims software has a lot of automation but very little real AI surface area. A lot of the things AI is good at (text generation) must have certain verbiage or use legally approved templates.
Granted, this may not be the case for all verticals, but this is just my observation from 20 years. (Mostly in smaller IA firms and state-based institutions).
Also, my wife has pointed out that there haven’t been any major cats in the last few years. Katrina kept people working for years. A lack of big storms has also contributed to this.
> Anecdotally, many of them are aging out
Ageing and out and neither being replaced directly or reduce numbers coming in at the bottom/mid as others move up, is still a drop in people. If part of the reason there is less need to bring people in lower down is that a few are being expected to do the work of a few more due to automation via AI, then that is AI replacing them, just not quite as directly as “we need less people as AI makes each more efficient so there are redundancies”.
There isn't a whole lot of room for automation in much of this.
Underwriting on the other hand, that is a different story.
It was precisely because of AI that I had the time and language to fight a recent claims report for a totaled vehicle, and squeezed an additional 25% from them. And it was still not adequate payout.
Let claim adjusters and insurance get off their high horse until they start being the bane of existence and paying what they owe clients.
They are using the cameras to catch fraud (where claimants try to lump previous damage with damage covered by the current claim.), etc.
I.e. Dentists are scammers at such high rates that their opinions on how to organize society should be actively ignored. The opposite or "rule by guild" is basically https://en.wikipedia.org/wiki/Syndicalism which is the alternative "terrible ideology" to Socialism that would have taken off in a world where Germany won WW2