The gap between 22% and 4% is the honest summary of agentic AI in insurance right now: widely intended, barely started.
Short answer: Industry reporting puts 22% of insurers expecting an agentic AI solution in production by the end of 2026 while only 4% had begun that move when surveyed, with adoption forecast to rise from 14% to 70% by 2028. More than 20 US jurisdictions have adopted the NAIC Model Bulletin requiring a written responsible-AI programme.
Key takeaways
- Industry coverage reports 22% of insurers expecting an agentic AI solution to be in production by the end of 2026, while only 4% had started that move when surveyed, with adoption forecast.
- 26 live listings in this category publish a rate, at a median top-of-range of $92 an hour and a ceiling of $175.
- The work is remote contract work, asynchronous, with no set hours and no guaranteed volume.
- Applications screen on a short skills assessment rather than a resume or interview.
What was reported
The finding
Industry coverage reports 22% of insurers expecting an agentic AI solution to be in production by the end of 2026, while only 4% had started that move when surveyed, with adoption forecast to rise from 14% currently to 70% by 2028. Vendor-published case figures cite a reinsurer cutting underwriting audit time 45%, from 200 hours to 110 hours per managing general agent. One platform reports governing underwriting decisions across more than $75 billion in annual commercial property and casualty premium for carriers including Aviva, Allianz, Sompo, Beazley and Markel. On governance, the NAIC adopted its Model Bulletin on the Use of AI by Insurers in December 2023, and by mid-2026 more than 20 US jurisdictions had adopted it, expecting insurers to maintain a written programme for responsible AI use covering transparency, testing and documentation. The performance figures above come from software vendors rather than independent research.
Reporting puts 22% of insurers expecting an agentic solution in production by the end of 2026 against 4% who had begun when surveyed, with adoption forecast to move from 14% to 70% by 2028. Treat the vendor-published performance figures around this, such as a reinsurer cutting audit time from 200 to 110 hours per MGA, as marketing rather than measurement.
What the listings pay
The governance side is better attested and more consequential. The NAIC Model Bulletin on the Use of AI by Insurers has been adopted by more than 20 US jurisdictions, and it expects a written programme covering transparency, testing and documentation.
| # | Role | Advertised rate | Platform |
|---|---|---|---|
| 1 | CAM Programming Expert (Fusion 360) | $120 to $175 an hour | Mercor |
| 2 | Operations / Program Management Experts | $80 to $160 an hour | Mercor |
| 3 | CAM Programmer | $80 to $120 an hour | micro1 |
| 4 | Infusions Insurance Verification Call Script Writer | $60 to $115 an hour | Mercor |
| 5 | Risk-adjustment / HCC coding leader | $110 to $110 an hour | Elion Health |
| 6 | E&S Insurance Underwriter - AI Trainer | $90 to $110 an hour | Mercor |
| 7 | Insurance Verification & Benefit Manager | $105 to $105 an hour | Elion Health |
| 8 | Software Expert (Development & Programming) | $100 to $100 an hour | Mercor |
| 9 | Insurance Domain Expert | $60 to $100 an hour | Mercor |
| 10 | GPU Programming Software Engineer | $60 to $95 an hour | micro1 |
Source: 26 live listings on this board that publish a rate, read directly from each posting on 2026-09-06. Listings without a published rate are excluded rather than estimated.
How this compares across the board
A rate only means something next to the alternatives. This is every category we track with at least five listings publishing a rate, ranked by median top-of-range, so you can see where this work sits rather than taking a single number on trust.
| Category | Listings | Median low | Median top | Highest |
|---|---|---|---|---|
| Legal | 95 | $100 | $140 | $400 |
| Medical | 68 | $77 | $120 | $400 |
| Consulting | 45 | $80 | $120 | $280 |
| Finance | 94 | $80 | $110 | $280 |
| Engineering | 114 | $70 | $100 | $300 |
| Research/PhD | 132 | $70 | $90 | $280 |
| Writing | 36 | $40 | $80 | $280 |
| Bilingual | 78 | $44 | $52 | $120 |
| Annotation | 25 | $12 | $24 | $120 |
Same source and date as above. Categories are matched on listing title, so a role can appear in more than one.
What it means for you
That is a regulatory requirement for people who can evidence how a model was tested, which is the same capability labs buy. Across 26 live insurance, reinsurance and actuarial listings on our board that publish a rate, the median top-of-range is $92 an hour, reaching $175.
Why reinsurance is slower than primary insurance
Reinsurance deals in low-frequency, high-severity exposure, which is the worst possible data profile for a model. The events that matter most are the ones with the fewest precedents.
Treaty structures are also bespoke. Program business, facultative placements and layered towers are negotiated arrangements rather than standard products, so there is far less repeatable pattern for a system to learn from.
That is why the automation is landing on audit, submission intake and document handling rather than on the risk judgment itself, which matches the pattern across every regulated category we track.
What the NAIC bulletin means in practice
A written responsible-AI programme covering transparency, testing and documentation is an evidentiary burden, not a technical one. Someone has to be able to show a regulator what was tested and why the testing was adequate.
That work resembles AI evaluation closely: you are asked whether an output is defensible and required to write down the reasoning clearly enough for someone else to act on.
Actuarial listings sit at the top of this category because they pair the domain with a quantitative credential. Underwriting, claims and program business roles ask for practical experience instead.
Who should apply
Two checks before you spend time on an application. Confirm the role accepts applicants from your country with the eligibility checker, since a meaningful share of listings carry location requirements. Then run the advertised rate through the take-home calculator, because this is contract work and the headline figure is before self-employment tax.
Applications complete on the hiring platform and usually take a few minutes, with a short skills assessment in place of an interview. Fill in every credential, language and professional background field on your profile. Those are what route you to the better paid listings, and most applicants leave them blank.
Frequently asked questions
How many insurers actually use agentic AI?
Reporting puts 22% expecting a solution in production by the end of 2026 while only 4% had started when surveyed, with adoption forecast to move from 14% to 70% by 2028.
How reliable are the efficiency figures?
Case figures such as cutting underwriting audit time from 200 to 110 hours per MGA come from software vendors rather than independent research, so treat them as marketing claims.
What governance applies?
The NAIC Model Bulletin on the Use of AI by Insurers, adopted by more than 20 US jurisdictions by mid-2026, expects a written responsible-AI programme covering transparency, testing and documentation.
Why is reinsurance adopting more slowly?
Low-frequency, high-severity exposure gives models the fewest precedents where they matter most, and treaty structures are bespoke rather than standard products.
What does insurance AI work pay?
Across 26 live insurance and actuarial listings publishing a rate, the median top-of-range is $92 an hour, reaching $175.
Do I need an actuarial qualification?
Only for actuarial listings, which pay most. Underwriting, claims and program business roles ask for practical experience instead.
What is program business?
Specialty insurance written for a defined class through a managing general agent, typically with bespoke terms, which is part of why it resists standardisation.
Sources
See every live role
The full board updates several times a week, with the advertised rate on each listing and closed roles removed.