04.22.26 By Bridgenext Think Tank
Underwriting has always been a judgment-driven discipline. That hasn’t changed. What has changed is the sheer volume of thinking underwriters must do before they can even apply that judgment. They’re navigating fragmented submissions, shifting risk signals, tighter regulations, and escalating loss volatility. The pressure is real: global insured catastrophe losses reached $80 billion in the first half of 2025, nearly double the 10-year average, continuing into 2026 with annual losses consistently exceeding $100 billion.1 In this environment, speed, clarity, and confidence aren’t just nice-to-haves; they’re essential.
Underwriting remains one of the most critical drivers of profitability, with performance improving in 72% of global markets in 2024, even as margin pressures intensify.2 That makes the real challenge clear: it isn’t access to information; it’s the ability to interpret it fast enough to make sound decisions. This is where AI copilots make their mark. They’re not here to replace underwriting expertise. They’re here to lighten the cognitive load that’s slowing it down.
This blog explores the growing strain on underwriting, how it shows up in everyday work, and how AI copilots can help teams move faster without compromising judgment.
Underwriting is becoming more challenging as underwriters must process data from more sources, assess rapidly evolving risks, and navigate greater regulatory scrutiny, all within the same limited timeframe. Data is pouring in from an ever-expanding array of sources, while 76% of US insurers have already implemented generative AI in at least one business function, adding even more inputs, signals, and expectations into underwriting workflows.3 At the same time, risks are evolving faster. Climate-related losses have consistently crossed $100 billion annually in recent years, making risk assessment more volatile and harder to predict.4
Emerging exposures like climate and cyber are not just adding complexity, they are reshaping underwriting itself. At the same time, regulators are increasing scrutiny around governance, transparency, and model explainability, especially as AI adoption accelerates across the industry. Clients, meanwhile, expect faster responses and a seamless experience.
All of this creates real pressure. Underwriters are now evaluating more variables, in less time, with greater accountability. The rhythm of their work has changed and the result is a cognitive bottleneck, one that directly impacts decision quality, profitability, and competitiveness.
Cognitive overload emerges when underwriters spend more time on repetitive, process-driven tasks than on actual risk analysis in their day-to-day work. They are often tied up with activities such as:
While individually these tasks are necessary, collectively they can be exhausting and the impact is real. In many cases, underwriting still takes 3–5 days for a decision, largely due to manual effort and fragmented workflows. With AI, that can drop to
minutes with over 99% accuracy, showing just how much time is lost in the current process.5 This friction shows up as:
| Metric | Manual / Fragmented Workflow | AI-Copilot-Assisted Workflow |
|---|---|---|
| Time to underwriting decision | 3–5 days5 | Minutes, 99%+ accuracy5 |
| Manual underwriting tasks | Up to 50% of an underwriter’s time, on average10 | Up to 67% reduction in manual task volume5 |
| Underwriting cycle time | Multiple days per submission on average, longer for commercial/specialty lines | 30%+ faster on average; some commercial P&C carriers report 60–99% quote-to-bind reductions10 |
| Information-gathering time (case example) | Manual, cross-system search for each query | ~5 minutes saved per query, on average (65,000 minutes saved across 13,000+ queries)5 |
| Query volume handled by one insurer’s AI assistant | — | 13,000+ queries post-rollout5 |
The trade-off is clear: every hour spent gathering information is an hour not spent making decisions. When underwriters can focus more on judgement than data collection, outcomes improve. At scale, the impact compounds. AI can reduce manual underwriting tasks by up to 67% and shorten cycle times by more than 30%, returning valuable capacity to the work that drives results: informed decision-making.
AI copilots matter now because insurers are under pressure to keep pace with growing underwriting complexity while maintaining speed, consistency, and profitability without adding headcount. They summarize submissions, retrieve relevant guidance, surface anomalies, and recommend next steps, removing the clutter around decisions so underwriters can focus on making them with clarity and confidence.
This is critical in a market where speed, accuracy, and consistency directly impact performance. The industry has already moved beyond experimentation. While many insurers are adopting generative AI in business functions, only a fraction of initiatives progress beyond proof of concept. This highlights a key insight: the tool itself isn’t enough; workflow integration is equally vital.
A copilot that operates outside the process adds another system to manage. A copilot embedded within the process becomes an integral part of how work gets done.
No. AI copilots are designed to support underwriting judgment, not replace it. AI is good at speed. It can process large volumes of information, spot patterns across thousands of data points, and flag potential risks in seconds. But underwriting decisions don’t live in clean datasets. They live in context, and that’s where human judgment comes in. Across the industry, even as AI adoption grows, insurers are not taking humans out of the loop. In fact, recent studies show that most AI systems in insurance are still designed with human oversight built in, especially in underwriting and pricing decisions.6
There’s a reason for that. Underwriters don’t just assess risk. They interpret incomplete information, balance trade-offs, negotiate with brokers, and make calls where there is no obvious “right” answer. This is especially true in commercial and specialty lines, where every risk has its own story. Regulation reinforces this reality. AI systems used in underwriting are increasingly classified as “high-risk,” requiring explainability, strong governance, and clear human oversight.7 Trust plays a role too. Nearly 1 in 4 consumers (24%) say they don’t fully trust how insurers use their data, which makes human accountability even more important in decision-making.8
So while AI can accelerate analysis, it cannot own the outcome. Underwriting still depends on knowing when to say yes, when to pause, and when to walk away. It’s about judgment, not just calculation. AI copilots support that process. They take away the repetitive load, surface what matters, and give underwriters the space to think more clearly. The decision, however, still belongs to the human.
In an AI-augmented future, underwriters spend less time gathering and validating information and more time applying judgement, managing portfolios, and advising. At the same time, AI handles the heavy lifting of data synthesis and retrieval. This shift delivers tangible operational efficiency and business value. It enables insurers to move faster, select risks with greater precision, price more competitively, and respond more agilely to market changes. It also elevates the insurer’s external presence. Brokers, clients, and investors see a company that is more responsive, capable, and future-ready. The opportunity is significant. Industry estimates show that AI-enabled underwriting can improve loss ratios, reduce expense ratios, boost productivity, and accelerate quote turnaround times. These aren’t incremental gains, they redefine how insurers compete.
AI copilots are only as effective as the environment in which they operate. Insurers that realize meaningful business outcomes build a foundational GrowthOS that connects data, embeds intelligence into workflows, and preserves human judgment where it matters most. That foundation rests on three core elements:
Each of these elements lands differently depending on where you sit in the organization:
| Role | What to Prioritize |
|---|---|
| CIOs / CTOs | Replacing legacy systems and siloed data that block real-time context |
| Heads of Data & AI | Moving from secure data storage to actionable, decision-ready data |
| Operations Leaders | Removing friction and reducing cost per decision |
| Business Leaders | Tracking measurable improvements in speed, quality, and consistency |
Underwriting is facing a cognitive overload. That’s the real challenge. AI copilots matter because they alleviate this pressure. They cut through the noise, enhance decision-making, and give underwriters the freedom to focus on judgment, relationships, and risk quality. But technology alone isn’t the solution. The true advantage lies in creating a connected underwriting ecosystem where people, data, platforms, and governance seamlessly align.
This is where Bridgenext makes the difference. By removing system friction and embedding intelligence into underwriting workflows, insurers can transform cognitive overload into faster, higher-quality decisions. The outcome? Superior risk results, scalable growth, and a commanding market position.
Connect with us to turn your working setup into a winning system.
It occurs when fragmented data, submissions, guidelines, and compliance tasks leave underwriters with less time for risk judgment.
An AI copilot summarizes submissions, retrieves guidelines, flags issues, and supports underwriters without taking over the final decision.
Automation executes predefined tasks, while an AI copilot works alongside underwriters to support real-time decisions and judgment.
No. They reduce repetitive work while keeping risk decisions and accountability with human underwriters.
Deployments have reduced 3–5 day decisions to minutes and cut manual tasks by up to 67%.
They often fail because the AI sits outside the workflow instead of being embedded into how underwriters already work.
References
1 www.oecd.org/en/publications/global-insurance-market-trends-2025_0d11ecf4-en.html
2 www.gartner.com/en/documents/5409263
3 www.gartner.com/en/documents/6678434
4 www.insurancebusinessmag.com/reinsurance/news/breaking-news/ai-raises-the-bar-for-reinsurance-underwriting–without-replacing-judgment-567980.aspx
5 www.insurancebusinessmag.com/us/news/technology/only-30-of-insurer-ai-projects-make-it-past-pilot-stage-report-finds-564367.aspx
6 www.guidewire.com/resources/blog/technology/how-ai-and-ecosystem-innovation-are-transforming-underwriting
