AI is changing commercial insurance underwriting by taking over the work around each decision: reading submissions, pulling third-party data, and checking classification against a carrier's appetite. Underwriters still write the rules and decide the exceptions. Adoption is uneven, and regulators now expect insurers to govern every automated outcome.
For an independent agent, the change shows up as fewer repeated questions and faster answers on routine files. Here's what changes at each stage, what stays with people, and why small commercial feels it first.
Key Takeaways
AI targets the non-judgment work around a submission: intake, data gathering, and classification checks.
Written underwriting rules decide which risks a person reviews, so the rules matter more than the software.
Regulators expect a written AI program and clear accountability, and the insurer stays responsible for every outcome.
What Is AI Underwriting in Commercial Insurance?
AI underwriting uses software to read submissions, pull third-party data, and check classification and eligibility against a carrier's appetite. The software also applies the carrier's written rating and selection rules. People write those rules and decide every exception.
Adoption is real but uneven. AM Best surveyed roughly 150 rated carriers and MGAs, and only about one in five described their AI implementation as advanced. Deloitte reports that 82% of carriers are planning agentic AI adoption within three years.
How AI Underwriting Works, Stage by Stage
Submission intake. The system reads the application and documents and structures them into one file.
Data gathering and pre-fill. It pulls third-party data and fills routine details.
Classification and eligibility checks. It compares the class, location, and exposures with the carrier's appetite.
Rating and decision. It applies the carrier's written rules to the finished file.
After the fourth stage, a risk takes one of three paths:
Routine risks that fit the written rules can move without manual review. The industry calls this straight-through processing (STP).
Borderline risks reach an underwriter with the research already assembled.
Unusual requests, such as re-classing a business or raising a limit, go to a person from the start.
That makes the written rules the most important part of the system. Each rule carries an underwriter's judgment forward to every file it touches.
Why Small Commercial Feels the Change First
A small account carries nearly the same file work as a larger one. Those costs are largely fixed per policy, so the expense ratio rises as premium falls. See what makes small commercial insurance unprofitable for traditional carriers and why small commercial insurance is so hard to place quickly.
Take a two-chair dental office moving into a new suite. You enter basic business details, and pre-fill supplies routine facts such as construction and occupancy. The class and location fit appetite, so the file moves on written rules.
Then the lease requires a higher liability limit than the standard offer. That request routes to an underwriter, who makes the call. Through both steps, you advise the dentist on what the lease demands.
What Stays Human, and How Regulators See It
The underwriter writes the rules and owns every exception. You own the client relationship and the coverage advice.
The NAIC issued its AI Model Bulletin in December 2023, and at least 24 states had adopted it by March 2025. It expects a written AI program, with senior management accountable to the board.
New York's Department of Financial Services issued Circular Letter No. 7 in July 2024, and insurers can't rely solely on a vendor's claim of non-discrimination. Accenture research reported by Insurance Journal finds regulators increasing their scrutiny of algorithmic bias. Responsibility for every automated outcome stays with the insurer, which is why the three routing paths matter.
What Underwriting Looks Like When AI Is Built In From the Start
Most carriers add AI tools to systems designed years ago for larger accounts, so the file work on a small submission shrinks less than you'd hope.
MGT Insurance, the first AI-native neo-insurer for small commercial property and casualty, built from the ground up for independent agents, was built around the stage-by-stage split. The neo-insurer model is explained in full elsewhere. The platform gathers third-party data, pre-fills routine details, and checks classification against appetite. A real underwriter stays in the loop for judgment calls, such as re-classing a business or raising a limit, and underwriting decisions stay with MGT. For how that holds at volume, see how AI-native carriers keep underwriting discipline while moving fast.
Aimee, MGT's quoting and appetite agent, gives a real-time read on whether a specific risk is in appetite. MGT is also the first carrier to offer chat-based quoting tools for independent agents, and every quote opens in the MGT portal for the agent to review before binding. The Business Owner's Policy is MGT's flagship line, written for medical offices like the dental practice above.
The fair question about any carrier built this way is whether it will be there when a claim comes. MGT holds an A- ("Excellent") financial strength rating from AM Best (AM Best, March 2025).
For what this means for a single quote on your desk, see how AI underwriting cuts quote time for independent agents.
Ready to Place Small Commercial With an AI-Native Carrier?
Already partnered with MGT? Start a quote and place coverage for your next small business client.
New to MGT? Partner with us to add a carrier built for independent agents, with fast, accurate quoting and broad class appetite.
Business owner? Work with an independent agent. MGT distributes exclusively through independent agents.
Final Takeaway
The real measure of AI underwriting is how much judgment it frees. Judge any market on how it handles the file that needs a person.
AI Underwriting FAQ
How is AI being used in small business insurance underwriting?
AI reads submissions, pulls third-party data, checks classification against appetite, and applies a carrier's written rules. Underwriters decide the exceptions.
What are the benefits of AI underwriting for small business policies?
Routine files move with fewer repeated questions, and classification gets checked before the quote rather than after a claim. Underwriters spend their time on the files that need judgment instead of on data entry.
Will commercial insurance underwriters be replaced by AI?
No. Underwriters set the rules and decide every exception, and agents keep the client relationship and the coverage advice.
Is AI underwriting regulated?
Yes. States that adopt the NAIC Model Bulletin expect a written, board-accountable AI program, and New York's Circular Letter No. 7 covers AI and external data in underwriting.
How does MGT use AI in underwriting?
MGT's platform gathers third-party data, pre-fills routine details, and checks classification against appetite. A real underwriter stays in the loop for judgment calls, such as re-classing a business or raising a limit, and underwriting decisions stay with MGT.
This content is for informational purposes only and does not guarantee coverage under any insurance policy. Actual coverage, terms, and exclusions are governed by the specific policy issued and may vary by state, carrier, and individual circumstances. Please review your policy documents or consult with your agent for guidance specific to your situation.
