Insurance by Algorithm: How AI Is Quietly Repricing Commercial Risk
If you work in commercial lines in any capacity, you have noticed something shifting in the carrier relationship over the last two years. Submissions that used to generate real conversations with underwriters are coming back faster, with less dialogue. Quotes appear that no one on the carrier side seems able to fully explain. Declinations arrive with language that sounds reasonable but does not actually tell you anything. Renewals move in directions that do not match the account's loss history.
The reason is not that underwriters have gotten lazier or that carriers have gotten more bureaucratic. The reason is that an increasing number of carriers have embedded artificial intelligence into the front end of the underwriting workflow. Not at the final approval stage. Not at the pricing sign-off. At intake. Before a human opens your submission.
This is not a vendor pitch or a technology forecast. More than three-quarters of U.S. insurers reported implementing generative AI across underwriting, claims triage, and customer engagement as of 2024. BCG estimates that 36 percent of total AI value for insurance can be captured across the underwriting function alone. The global market for AI in insurance stood at approximately $15 billion in 2025 and is projected to reach $246 billion by 2035.
Those numbers represent a structural shift in how risk gets assessed, how appetite gets communicated, and how pricing gets set. For retail and wholesale brokers, the shift carries consequences that go well beyond efficiency. It changes what submission quality actually means. It changes what carrier relationships are worth. And it changes the definition of what a skilled placement professional does for a living.
So let's break this down. What AI underwriting actually does inside carrier systems. How it is changing pricing outcomes in commercial property, casualty, E&S, and cyber. Why the transparency problem is real and not just a regulatory talking point. And what brokers need to change right now in how they operate.
What AI Underwriting Actually Does
The word AI gets used loosely in insurance, often to describe anything from a simple rules-based filter to a sophisticated neural network. In underwriting, the meaningful distinction is between tools that speed up tasks a human was already doing and tools that make decisions a human used to make. Both exist in the market today, and understanding the difference matters because they create different problems for brokers.
At the most foundational level, AI has entered submission ingestion. Large language models now handle document processing, entity extraction, and risk profile construction for incoming submissions. What used to be a manual review of loss runs, financial statements, and application data has become an automated process that extracts key variables, identifies missing information, and generates a structured risk assessment in minutes. According to AWS's financial services division, this kind of intelligent document processing is cutting manual processing times from days to minutes at multiple carriers.
From that structured risk profile, predictive models assess claims frequency and severity against the carrier's portfolio. These models pull from historical loss data, third-party enrichment sources, satellite imagery, property characteristics databases, and, in commercial lines, real-time business intelligence including financial benchmarks, litigation exposure data, and operational signals drawn from publicly available sources. One of the most significant applications now deployed at several carriers involves natural language processing of websites, social media presence, and public records to build dynamic business risk profiles that flag contractor violations, regulatory exposure, or financial distress before a human underwriter has reviewed a single page.
For property lines specifically, computer vision has added a data layer that brokers have historically controlled. Aerial and satellite imagery processed through AI models now give carriers independent visibility into roof conditions, fire exposure proximity, flood plain adjacency, and construction quality without depending on what appears in the submission. Cape Analytics, backed by roughly $75 million in funding, has built a platform that allows commercial property portfolios to be underwritten based on actual observed physical condition rather than age or replacement cost estimates. AIG reported a 15 percent improvement in underwriting data collection and accuracy rates after deploying generative AI platforms in early 2025.
Carriers are also using AI for submission triage at scale. Platforms like Sixfold allow E&S carriers to sort through high submission volumes instantly, flagging accounts that align with their appetite and deprioritizing those that fall outside it before routing to the appropriate workflow. The partnership between Harbor.ai and Carpe Data operates the same way inside E&S underwriting workflows, surfacing appetite-aligned risk factors, location vulnerabilities, and regulatory concerns so underwriters can make decisions faster and more consistently.
What this creates in aggregate is an underwriting environment where your submission is being evaluated against standards that are invisible to you, applied before a human reviews it, and shaped by data you did not provide. The carrier's AI has already formed an opinion by the time your underwriter contact looks at the account. In some commodity lines, that AI opinion is the binding decision.
Gartner projects that by 2027, 60 percent of underwriting decisions will be shaped by adaptive AI models rather than hard-coded business rules. That is not far away. And the implications for how brokers approach market relationships, submission preparation, and pricing negotiations are enormous.
Commercial Property: Pricing Before the Submission Arrives
Commercial property has seen the most aggressive AI deployment of any line, driven by catastrophe modeling pressure, reinsurance cost increases, and the growing gap between what brokers submit and what carriers can independently verify.
The 2025 Los Angeles wildfires pushed catastrophe models and reinsurance layers to their limits. Insured natural catastrophe losses surpassed $150 billion in 2024 as reinsurers continued shifting toward higher attachment points, leaving more primary exposure with carriers. That pressure has accelerated adoption of AI tools that allow carriers to stop depending on broker-submitted information for property risk assessment and start building their own independent view of every account.
The practical consequence for retail and wholesale brokers is significant. When a carrier's AI system can independently assess a property's fire exposure proximity, roof condition, flood vulnerability, and construction quality from aerial imagery before your submission is reviewed, the information you provide in the ACORD application is no longer the primary input into the underwriting decision. It becomes a corroboration check. If what you submitted aligns with what the AI observed, the account moves forward efficiently. If it does not, it surfaces as a rate adjustment, a restrictive condition, or a declination with language that sounds reasonable but does not actually tell you what triggered it.
For commercial property submissions, this means data quality is no longer just about following submission best practices. It is about pre-empting the AI's independent assessment. Brokers who understand what third-party data sources carriers are pulling, and who proactively address discrepancies between the insured's representation and the external data record, are in a fundamentally stronger position than those still submitting ACORD applications the way they always have.
Machine learning models are also being used to predict claims frequency and severity across industries and geographies in ways that introduce new pricing dynamics for property-heavy accounts. A hospitality account in coastal Florida, a mid-size warehouse in a convective storm corridor, or a habitational portfolio with aging construction in a secondary market will encounter AI-driven pricing signals that reflect the carrier's portfolio-level accumulation concerns, not just the individual account's loss history. The account may have a clean loss run. The AI may still generate a restrictive quote based on the carrier's modeled exposure at that specific location.
Brokers who have built client relationships on the premise that a clean loss history drives favorable terms are finding that premise challenged in ways that are hard to explain. The honest answer is that the pricing is no longer purely about the account. It reflects a geospatial risk calculation the carrier's AI performs against its entire book of business. A broker who can articulate that dynamic, and who can help clients understand how to differentiate their risk presentation in ways that stand out against the carrier's modeled assumptions, brings demonstrably more value than one who simply shops the account and reports back what the market returned.
The E&S market has made this explicit. Multiple sources tracking the 2026 surplus lines outlook have been consistent on one point: underwriters now prioritize digitally processed submissions with quality data. Brokers who provide incomplete applications risk being deprioritized in favor of those who present a clear, well-documented account from the start. With more capacity entering certain lines, carriers have options. The AI helps them sort submissions before the underwriter allocates attention.
Casualty and General Liability: The Invisible Risk Score
Casualty AI applications are somewhat less visible than property technology, largely because the data inputs are more behavioral than geographic. But the impact on how casualty and GL accounts get priced and triaged is equally significant for brokers working in these lines.
In workers' compensation, predictive models now incorporate movement data, ergonomic risk assessments, and jobsite sensor information alongside traditional underwriting inputs. Carriers are building real-time risk profiles for commercial accounts in safety-sensitive industries by pulling from operational data that small and mid-size insureds often do not realize is observable. The granularity of risk assessment available to AI-equipped carriers has created a market where identical-looking accounts from a submission standpoint can receive materially different pricing based on observable operational differences the broker never addressed in the narrative.
In general liability, natural language processing is doing something that should concern any broker who has been submitting a straightforward ACORD on a habitational, construction, or professional services account and treating that as sufficient. These models analyze publicly available information about the insured's business including their website, social media presence, online reviews, court records, and regulatory filings to identify operational red flags that do not appear in the submission. A contractor whose website lists services that fall outside the described operations. A habitational operator with unresolved code violations in public records. A professional services firm with recent litigation exposure not disclosed in the loss history. The AI surfaces those signals before the underwriter evaluates the account.
This creates two implications that run in opposite directions. For brokers who represent well-run, transparent accounts, the AI's independent assessment becomes an ally. If the client's external footprint supports a favorable risk narrative, proactively including a business narrative that aligns with what the AI will find independently makes the submission more credible, not less. For brokers who have historically relied on minimal disclosure or who represent accounts with operational complexity that is hard to present cleanly, the AI is now a source of pricing pressure that cannot be managed through carrier relationship alone.
The casualty market has also seen meaningful development in AI-driven triage tools that prioritize submissions based on how closely they resemble the carrier's historical bind patterns. Platforms compare new submissions against previously bound risks and surface accounts that match profitable segments of the existing portfolio. For wholesale brokers in particular, this means submission quality and data completeness are becoming factors in how quickly an account gets reviewed, not just how favorably it gets priced. An incomplete submission on a complex GL account no longer just creates a back-and-forth clarification cycle. It can cause the account to deprioritize in the underwriting queue while cleaner submissions from competing operations move ahead.
The industry has been direct about this. The cost of submitting a messy, incomplete account used to be a time delay. Today it can be a deprioritization that effectively removes the account from contention before a human underwriter allocates meaningful review time.
E&S and Specialty: Where AI Is Accelerating and Complicating Placement
The E&S market surpassed $81 billion in premiums in 2024, up 12 percent year over year, and has expanded at roughly 14 percent annually over the past decade. That growth has attracted significant AI investment from both carriers and distribution technology platforms, creating a segment where the placement workflow is being reshaped faster than most brokers have registered.
The challenge in E&S has always been speed and fit. E&S moves fast, risks are non-standard, and brokers shop across multiple carriers simultaneously. A slow response time means lost opportunities to a competitor who quoted faster. AI systems that can instantly triage incoming submissions against a carrier's appetite, surface the most relevant risks for underwriter attention, and deprioritize those that fall outside the risk profile have made E&S underwriting substantially more efficient for carriers. The question for wholesale brokers is how to operate in a world where the carrier's first evaluation of your submission is automated.
In February 2026, MGT, an AI-native carrier, announced a partnership with Amwins, the largest independent wholesale distributor in the United States, to bring AI-driven underwriting and pricing to select E&S risks. The collaboration combines Amwins' reach across 138-plus offices globally with MGT's platform that underwrites complex commercial risks instantly rather than in days. This is not an isolated development. Harbor.ai and Carpe Data's partnership specifically targets E&S insurers with AI tools that allow faster segmentation and pricing while reducing the manual research burden on underwriters evaluating complex submissions.
The strategic implication for wholesale brokers is nuanced. AI-accelerated E&S underwriting benefits brokers who operate efficiently and submit clean, well-organized accounts. Faster turnaround times mean better service delivery to retail partners. But it also creates a selection mechanism where submissions that do not meet data quality thresholds are automatically disadvantaged before a human advocate for the account can make the case.
The 2026 E&S market forecast also highlights a dynamic that is reshaping the risk classes flowing into surplus lines. Emerging exposures including AI liability, cryptocurrency custody, autonomous vehicle operations, and digital asset risks lack the actuarial history needed for conventional pricing models. These accounts belong in E&S precisely because standardized answers do not exist yet. But the same AI tools that are making standard E&S placement more efficient are still developing their frameworks for novel exposures.
Here is the thing nobody wants to say about this: for wholesale brokers operating in specialty niches, being ahead of the AI's understanding of a risk class creates differentiated value that is genuinely hard to replicate. A wholesale broker who can build a submission narrative for an AI liability exposure that speaks to carrier underwriting logic, acknowledges data limitations, and frames the risk within a coherent structure will consistently out-execute competitors who submit these accounts as if they were conventional placements looking for any available market. The absence of AI-friendly submission infrastructure for emerging risks is, paradoxically, where human expertise creates the most sustainable competitive advantage right now.
Wholesale brokers who engage underwriters beyond the transaction, sharing market intelligence and account context that feeds into the carrier's evolving risk model, are building relationships that strengthen their position in an AI-mediated placement environment rather than being displaced by it.
Cyber: The Line Where AI Is Both the Risk and the Underwriter
Cyber insurance sits in an unusual position in this discussion because artificial intelligence is simultaneously the source of escalating risk and the primary tool carriers are using to assess and price it. Understanding both sides of that dynamic is essential for any broker placing cyber coverage.
On the risk creation side, AI has dramatically lowered the barrier to entry for sophisticated cyber attacks. Deepfake-enabled social engineering, AI-accelerated phishing campaigns, automated vulnerability scanning, and adversarial attacks on AI models themselves represent a new generation of exposures that did not exist at meaningful scale two years ago. Coalition has already built AI-specific cyber coverage endorsements addressing deepfake fraud directly. Munich Re's aiSure product suite brings reinsurance backing to AI-specific risk structures. The emergence of affirmative coverage for AI-related exposures, covering software malfunctions, algorithmic biases, and cybersecurity vulnerabilities in AI-driven systems, signals that the market is beginning to price a risk category that will only grow from here.
On the underwriting side, carriers are deploying AI to monitor risk in ways that extend well beyond the initial transaction. Submission processes, real-time monitoring tools, and continuous risk signals are becoming increasingly automated. Insurance Business Magazine reported recently that expanded AI-based monitoring over the policy term is expected to become standard practice, meaning the underwriting evaluation does not end at bind. The carrier's view of your client's risk is no longer static.
The cyber underwriting AI environment creates a specific challenge around transparency. Unlike a human underwriter who can explain a rate increase in terms of an industry class trend or a specific claim concern, an AI-driven pricing adjustment may reflect a combination of signals that is genuinely difficult to decompose. Third-party risk management controls, network architecture signals, software patching behavior, employee security training documentation, and claims history all feed into the model in ways that are not always surfaced clearly in the underwriting communication. Woodruff Sawyer flagged in their 2025 Cyber Looking Ahead Guide that underwriting scrutiny around third-party risk management controls would intensify significantly, a trend that has continued into this year.
The broker's role in cyber has become partly technical and partly interpretive as a result. On the technical side, a broker who understands what security controls cyber AI models weight most heavily, and who can coach clients through implementing those controls before submission, has a measurable impact on pricing outcomes. On the interpretive side, a broker who can translate an AI-influenced coverage decision into terms the client understands, including explaining why a deepfake exclusion appeared mid-renewal or why premium moved despite a clean loss history, is delivering advisory value that clients cannot replicate by going direct.
WTW's analysis of emerging AI exposures and cyber coverage gaps makes a point every cyber broker should internalize: gaps and grey areas in coverage represent opportunities to protect clients further. The AI-driven risk creation environment is moving faster than policy language is updating. Brokers who identify these gaps proactively, rather than waiting for a claim to expose them, are operating at a level that justifies the premium they receive for their services.
The Black Box Problem and Why Brokers Cannot Ignore It
The transparency deficit in AI underwriting is not a peripheral concern. It is a structural problem that brokers will encounter directly in client conversations, in carrier negotiations, and in regulatory environments that are beginning to demand accountability.
Here is the core issue: when an AI model generates a pricing recommendation, issues a declination, or applies a restrictive endorsement, it does so through a process that is often not explainable in plain terms even to the carrier professionals who oversee it. The model has identified a pattern in data that produces a risk signal. But the specific combination of variables that triggered the output may be spread across dozens of input fields, enriched by third-party data the broker never saw, and weighted in ways that cannot be readily communicated to the insured or their representative.
Regulators have noticed. The NAIC released its model bulletin on AI in insurance in December 2023, and as of August 2025, 23 states and Washington D.C. had adopted it while four states had implemented specific AI regulations. The New York Department of Financial Services enacted Circular Letter No. 7 in July 2024, requiring carriers to establish internal governance frameworks and develop clear, simple explanations of how AI factors into underwriting decisions. California and Colorado have enacted regulations requiring carriers to justify risk decisions made with algorithmic input, particularly when those decisions result in adverse outcomes for the insured.
For brokers, the regulatory pressure on carriers creates both an obligation and a practical tool. The obligation is to understand, at a functional level, how a carrier's AI systems are influencing the terms your client receives. When a client asks why their premium increased 18 percent at renewal on an account with no losses, and the honest answer is that an AI model identified a change in their exposure profile based on satellite imagery and public business data, the broker who can frame that explanation clearly is far more valuable than one who can only say the market hardened.
The practical tool is this: in states that have adopted the NAIC model bulletin or enacted specific AI regulations, the carrier is obligated to have documented the decision logic. A broker who knows this, and who knows how to invoke the regulatory framework appropriately, can sometimes surface information that leads to a different outcome on an account that was initially scored unfavorably.
The bigger opportunity, though, is in the space between what AI can do and what it can understand. AI systems can process extraordinary volumes of data and identify statistical patterns with precision no human team can match. They cannot evaluate context, nuance, or the kind of operational story that transforms a borderline submission into a favorable placement. A client who has invested in loss prevention infrastructure, implemented safety culture changes not yet reflected in historical loss data, or undergone a management transition that materially improved the risk profile has a story that belongs in the submission narrative. The broker who knows how to tell that story, and who knows which elements of it give the underwriter a basis to advocate for the account above what the AI initially scored, is performing a function that AI cannot currently replicate.
That function is exactly where the value of the professional broker lives right now.
What Brokers Need to Change: Retail and Wholesale Perspectives
What Retail Brokers Need to Do Differently
The most immediate change required of retail brokers is in how they think about submission preparation. The narrative section has always been optional in the sense that it is not required to generate a quote. In an AI-mediated underwriting environment, the narrative is where the broker's value actually lives. It is where you provide context the AI cannot access independently, address discrepancies between the insured's self-representation and the data the carrier will pull from third-party sources, and frame the risk in terms that give the human underwriter a basis to override or refine the AI's initial scoring.
This means retail brokers need to know their accounts better than they currently do. Understanding what a carrier's AI might find when it runs independent data enrichment on your client requires knowing the client's business deeply enough to anticipate the questions. Does the website describe operations that are broader than the scheduled classification? Are there public litigation records that are not in the disclosed loss history? Has the business expanded into new service lines or geographies not updated in the policy record? These are the gaps AI surfaces, and which human underwriters escalate into pricing conversations. Retail brokers who address them proactively, in the submission rather than after the question, are operating at a higher level of service that clients feel even if they cannot always articulate why.
Retail brokers also need to develop more systematic relationships with wholesale partners around AI literacy. Wholesale brokers who track carrier AI deployment, understand which E&S carriers are using triage systems and how those systems are weighted, and share that intelligence with retail partners are providing services that go well beyond market access. Finding wholesale partners who operate at that level of sophistication is a meaningful competitive advantage for retail brokers serving complex commercial accounts.
What Wholesale Brokers Need to Do Differently
Wholesale brokers face a different but related challenge. The efficiency gains that AI is creating on the carrier side are compressing the timeframes in which placement decisions get made. An E&S carrier using AI triage can process a submission and generate a preliminary disposition faster than the wholesale broker can follow up. In this environment, submission quality at first touch is more critical than it has ever been. The wholesale broker who submits incomplete accounts to multiple markets and waits to see what comes back is operating in a manner that the carrier's AI is actively designed to deprioritize.
Building AI-fluent submission infrastructure means understanding what data fields and narrative elements the carrier's triage system weights most heavily for each class of business. It means maintaining client data in forms that allow clean, complete submissions to be prepared rapidly rather than assembled from scratch for each renewal. And it means developing a clear market strategy before submission rather than broad-shopping an account and using carrier responses to figure out the placement strategy.
The Amwins and MGT partnership on E&S AI underwriting points toward a future in which distribution partners who can integrate with carrier AI platforms will have structural advantages in speed and data quality over those who cannot. Wholesale brokers evaluating technology investments should be asking not just about internal workflow efficiency but about how their platforms interact with carrier underwriting systems. The integration layer between wholesale submission infrastructure and carrier AI triage is where competitive advantage will increasingly be determined.
For complex specialty placements, the wholesale broker's role as a translator between client exposure and carrier appetite becomes more valuable as AI handles commodity work, not less. A wholesale broker who understands the technical parameters of an AI liability policy, who can explain coverage gaps to a retail broker placing a tech client, and who can negotiate bespoke endorsements for exposures that fall outside the standard AI-underwritten framework is providing value that no automation tool can replace. The wholesale market is differentiating, and the differentiation is happening between those who are specializing deeply and those who are staying broadly generalist.
The Human Underwriter Is Not Going Away
One of the concerns that circulates in carrier offices is whether AI deployment represents a threat to human underwriting roles. The data suggests the answer is nuanced, and brokers should understand the nuance because it affects how they engage with underwriting contacts.
Industry surveys from 2025 showed that fewer than half of underwriters and actuaries feared being replaced by AI, down from nearly three-quarters and four-fifths respectively just a year earlier. This shift reflects a more pragmatic understanding of where AI creates value and where it does not. Capgemini estimates that underwriters spend as much as 41 percent of their time on administrative and operational activities rather than actual underwriting judgment. As much as 60 percent of broker submissions are never reviewed at all, and only 25 percent of those that are proceed to written policies. AI is addressing the front-end intake and triage burden, freeing underwriters to focus on the accounts that actually belong in their decision queue.
What this means in practice is that human underwriters are being repositioned as decision-makers on complex accounts, as advocates within the carrier for favorable treatment on accounts that warrant it, and as relationship managers for broker and distribution partnerships the AI cannot handle. For brokers, that means the accounts where you need a human underwriter to go to bat for you are the ones where the AI model has flagged complexity or risk signals that standard triage would deprioritize. Your ability to provide a compelling, contextual narrative that gives the underwriter a basis to advocate within the carrier is exactly what moves an account from AI-flagged complexity to favorable placement.
The accounts that concern sophisticated market participants are not the small commercial accounts AI handles end-to-end. Those were never relationship business. The concern is in middle market and large commercial, particularly in specialty lines, where the AI is increasingly influential but the human underwriter still makes the final call. In those cases, the AI's initial scoring shapes the conversation the underwriter has with their own management about appetite and pricing authority. A broker who understands this dynamic and prepares submissions that address the AI's likely objections before the underwriter encounters them is operating with a structural advantage that compounds over time.
The Regulatory Pressure Building Behind All of This
The regulatory environment around AI in insurance is developing rapidly, and brokers who track it will find both compliance obligations and useful leverage in what state regulators are requiring of carriers.
The NAIC model bulletin provides a framework for insurer AI governance, but individual state implementation varies significantly. New York's Circular Letter No. 7 is among the most specific, requiring carriers to develop governance frameworks for AI oversight and to provide clear explanations of how AI factors into underwriting decisions. California and Colorado have enacted regulations requiring carriers to justify decisions made with algorithmic input, particularly when those decisions result in adverse outcomes.
The broader regulatory trajectory is toward more transparency and explainability requirements, not less. Carriers that built governance frameworks early are entering 2026 better positioned to satisfy regulators and to build trust with distribution partners and clients. The IAIS Global Insurance Market Report for 2025 explicitly identified AI model governance and transparency as supervisory priorities.
Brokers should be asking their carrier partners what governance frameworks they have in place for AI in underwriting. Not because the answer will change the placement decision today, but because it signals carrier maturity and reduces the risk of arbitrary or unexplained pricing outcomes down the line. And in the states that have already enacted specific AI regulations, understanding the framework well enough to invoke it on behalf of a client who received an adverse AI-influenced decision is becoming a real part of professional broker practice.
What Changes Right Now
The shift toward AI-mediated underwriting is not a future condition to prepare for. It is the current market. The following adjustments apply immediately.
Audit your submission process for the data gaps carrier AI systems are most likely to surface. For commercial property accounts, this means understanding what satellite and aerial imagery shows about your clients' locations and proactively addressing any discrepancy between the physical condition of properties and what the application represents. For casualty and GL accounts, it means reviewing publicly available information about your clients' operations for signals that could trigger adverse scoring before you send the first submission to market.
Build submission narratives that address AI-likely objections directly. A well-constructed narrative tells the underwriter what the AI model might have flagged and explains why it does not represent the risk story accurately. Management quality, loss prevention investments, operational improvements, and business changes that would not appear in historical data but that materially improve the risk profile belong in the submission narrative. They give the human underwriter a basis to advocate for the account within the carrier system.
Develop a working knowledge of which carriers are using AI triage in which lines and what those systems weight most heavily. This is market intelligence that wholesale partners, carrier relationships, and trade publications can help you build. It will inform your submission preparation, your market selection strategy, and how you counsel clients on presenting their risk profile in ways that align with carrier appetite.
Expand your value proposition around the transparency problem. Clients receiving AI-influenced pricing adjustments they do not understand need advisors who can explain what is happening and why. This is advisory work the carrier will not do for your client and that no digital platform replaces. The broker who positions themselves as the client's interpreter of an increasingly automated marketplace is building a service model that AI cannot commoditize.
Track the regulatory landscape in your operating states. Understanding which AI regulations apply to carriers writing your business tells you what rights you and your clients have when AI-influenced decisions produce outcomes that do not reflect the actual risk.
The Takeaway: The Broker Who Understands the Algorithm, will Win
The insurance market has always rewarded brokers who understood underwriting better than the underwriters expected them to. The broker who could look at an account the way a carrier thinks about portfolio risk, who could anticipate objections and build the case for favorable treatment before being asked, has always out-executed the broker who simply moved paper.
AI has raised the stakes on that skill set significantly. The carrier's view of your client's account is now shaped before you have a chance to tell the story. It is built from data you did not provide, weighted by models you cannot see, and delivered to the human underwriter as a preliminary disposition. Your job is to understand what that model likely produced and to give the underwriter something better to work with.
The brokers who are positioning themselves well share a few characteristics. They know their accounts deeply. They prepare submissions that address the AI's likely objections rather than assuming the account will sell itself. They have developed market intelligence about where AI is being deployed and how it is changing placement dynamics in their specialty areas. And they have invested in the kind of advisory relationship with clients that justifies their role as a trusted intermediary in a market that is becoming increasingly automated for everything transactional.
The algorithm already priced your account. The question is whether you put the information in front of the underwriter that changes what happens next.
Fabio Faschi is an Insurance leader, National Producer, Board Member of the Young Risk Professionals New York City chapter and Committee Chair at RISE with over a decade of experience in the insurance industry. He has built and scaled over a dozen national brokerages and SaaS-driven insurance platforms. Fabio's expertise has been featured in publications like Forbes, Consumer Affairs, Realtor.com, Apartment Therapy, SFGATE, Bankrate, and Lifehacker.
