Claims Adjusters, Examiners, and Investigators

On high-exposure property claims, there are a lot of issues that AI will not be able to address. Soft fraud or claim exaggerations, unique structures, with unique circumstances, that change the way a claim is handled, reviewed, and evaluated. I do not think AI will replace the adjusters on the large and complex claims.

I worked as a Benefit Claims Examiner at Pacific Blue Cross, handling Regular (level 1) claims, MISC (level 2) claims, and eClaims. To adjudicate claims, I used ACES (a data entry system), BlueHub (where we received claim forms, prior authorizations, receipts, invoices, doctor's notes, and customer notes), and eINQUIRY (where we updated member information like birthdays and addresses, and checked their claim and reimbursement history). In ACES, I entered the dollar amount, service code, quantity, and attached doctor's notes. ACES would then calculate and approve or reject the claim, providing reasons for rejections, though sometimes manual approval was required. The work was straightforward. Claims examiners at PBC are unionized, but some don't process claims correctly, make errors, or manipulate their work to improve KPIs. I believe the CEO and IT department could develop an advanced AI system to fully automate level 1 and 2 claims, including eClaims, saving the company money and improving customer satisfaction. Currently, eClaims and Dental claims are heavily backlogged, leading to complaints about incorrect processing and delayed reimbursements. I hope AI will eventually handle all these tasks, as many claims examiners are overpaid for subpar work, with the Union defending those who lack competence, especially long-time employees.

Takes some negotiation and people skills, but a large portion of the job is interpretation of policy documents which can probably be done very effectively by ai

Lots of human interaction, negotiation

Some insurance claims can not be evaluated solely by using rules. Intuition and human judgement are needed in evaluating some claims.

No, the fact that a computer cannot detect a lie is a problem. As some small lies someone tells might not effect the claim. I don’t think a computer will be smart enough to know the difference

"Claims adjuster" is a very broad term. I currently do field investigations of large commercial property losses, and there's quite a bit of nuance when determining if a claim is covered as well as the negotiations with different experts, contractors, and etc. I feel like automation may help with the job, but don't see the profession being removed completely.

not possible ever

Tech is already here - drones can do a full exterior inspection, use object detection for finding roofing damage, spit out an estimate better than 7/10 adjusters. Industry has been on a downward path since 2005, its only going to get worse. What's left is dealing with PAs, ex-convict roofers, Karens who freak out it'll take the drones 3 days to do an inspection, and OCD nonsense from carriers.

investigation is needed in this occupation