Will AI Replace Underwriters? The Numbers Behind the Shift

Insurance underwriting has always been about sizing up risk. You review an application, weigh the data, check the actuarial tables, and make a call — approve, deny, or modify the terms. It is a job built on pattern recognition and judgment, which is exactly why AI is making rapid inroads. Our data shows AI exposure for insurance underwriters at 64% in 2025, up from 52% just two years ago, with automation risk at 62/100.

Those are some of the highest numbers in the financial services sector. But before you update your resume, the full picture is more nuanced than the headline suggests.

Where AI Is Already Doing the Work

The clearest impact is in routine risk assessment. AI systems can now process standard applications — homeowners insurance, auto policies, straightforward commercial lines — faster and more consistently than human underwriters. These systems pull data from dozens of sources simultaneously, run it against historical loss patterns, and generate pricing recommendations in seconds rather than hours.

Predictive modeling has transformed how carriers evaluate risk. Machine learning algorithms can identify correlations in claims data that no human would spot, from the relationship between specific building materials and fire loss frequency to the subtle patterns that predict auto claims. A senior underwriter told us the models now catch risk factors that even experienced professionals miss.

Document processing is another area where automation is well advanced. AI can extract relevant information from applications, financial statements, inspection reports, and medical records, then flag inconsistencies or missing data. What used to take an underwriter thirty minutes of reading and data entry now happens in under a minute.

Portfolio monitoring has also shifted. AI systems continuously scan existing books of business for emerging risks — a manufacturing client that just received an OSHA citation, a commercial property in the path of changing weather patterns, a medical practice facing new malpractice trends. This real-time monitoring was simply impossible at scale before.

What Keeps Underwriters in the Game

Complex and unusual risks still need human judgment. When a tech startup wants coverage for a novel product, when a manufacturer is expanding into a country with limited loss data, or when a claim history shows an unusual pattern that could mean either bad luck or fraud, experienced underwriters bring something AI cannot replicate: the ability to weigh ambiguous information and make judgment calls that balance risk with business opportunity.

Relationship management is another anchor. Underwriters who work with brokers and agents are not just processing paper — they are building partnerships, negotiating terms, and making exceptions that make business sense. A broker who brings a borderline account needs a human who understands context, not an algorithm that says no.

Regulatory navigation matters more than ever. Insurance regulation varies dramatically by state and line of business, and the rules change constantly. Underwriters who understand the regulatory landscape can structure coverage in ways that meet both carrier guidelines and regulatory requirements, something AI systems struggle with given the complexity and constant evolution of insurance law.

The theoretical AI exposure sits at 87% — meaning the technology could potentially handle most underwriting tasks. But observed exposure is only 38%, reflecting the gap between what AI can theoretically do and what companies have actually implemented. That gap exists because of regulatory caution, integration challenges, and the genuine need for human oversight in consequential financial decisions.

The 2028 Outlook

Projections suggest AI exposure will reach roughly 72% by 2028, with automation risk climbing to 68/100. The trajectory is clear: routine personal lines underwriting will become almost entirely automated, and even standard commercial lines will see heavy AI involvement. The underwriters who thrive will be those handling complex risks, managing key broker relationships, and overseeing the AI systems that handle everything else.

Career Advice for Underwriters

Specialize in complex risk classes where human judgment remains essential — think emerging technologies, international exposures, or novel coverage structures. Develop your relationship skills with brokers and agents. Learn to work with AI tools rather than competing against them — the underwriter who can evaluate and override an AI recommendation with sound reasoning is far more valuable than one who simply duplicates what the machine already does. Consider the growing field of AI model governance in insurance, where underwriting expertise meets technology oversight.

For detailed automation data on this occupation, see the Insurance Underwriters page.

---

This analysis is AI-assisted, based on data from Anthropic's 2026 labor market report and related research.

Update History

• 2026-03-25: Initial publication with 2025 baseline data.

Related: What About Other Jobs?

AI is reshaping many professions:

• Will AI Replace Compliance analysts?
• Will AI Replace Fraud investigators?
• Will AI Replace Doctors?
• Will AI Replace Chefs?

Explore all 470+ occupation analyses on our blog.