The AI panic is in full swing. AI will replace jobs, upend business models, and destroy value. Software companies will die and many industries and professions will be rendered obsolete. Governments will need to implement universal basic income schemes just so humans can survive while machines do all the work.
Property/casualty insurance has not been immune with many predicting dire scenarios for the industry. The news that a Spanish insurer had a homeowners insurance app approved by OpenAI sent stocks of major brokers crashing. Others are predicting that the need for underwriters, adjusters, agents and other industry professionals will plummet as agentic AI takes on more complex tasks. These reactions reflect a massive misunderstanding of commercial lines insurance and the role of brokers.
Permit me to express a different view – perhaps a contrarian view –based on my participation in prior major tech waves and witnessing their implications for insurance, especially the Internet and the digital era.
The internet, direct distribution, and agents
In the 1990s at the dawn of the Internet for business, the widespread belief was that the role of agents would diminish; disintermediation was the byword, meaning that customers would be able to go directly to insurance carriers to purchase insurance. Fast forward to 2026 and we find the agency world thriving. In the US market, there are virtually the same number of independent agencies (IAs) today as there were in 2000. It’s true that the exclusive/captive agency force has declined from about 55,000 in 2000 to 35,000 in 2025. And yes, some of that is due to customers connecting directly online with insurers for their insurance (especially personal auto and small commercial). However, most of that decline is reflective of carriers shifting toward IA channels. An interesting data point is that 87% of commercial lines insurance in 2025 was sold through independent agencies – up from 83% ten years ago.
The digital era and P&C
Over the last 15 years, global digitalization and mobile tech have driven massive transformation in society and business, with the insurtech era harnessing these technologies to “disrupt” insurance.
Insurtech firms and the adoption of new technologies by existing P&C companies and incumbent tech providers have resulted in significant changes over that period, improving the customer experience, risk selection, claims, and operational efficiencies. Have the
advancements in the digital era fundamentally changed the business model of insurance? No. Have they dramatically reduced the number of underwriters, adjusters, and other professionals? No. Have they improved the productivity of these individuals? Absolutely yes. Have they created new opportunities for covering risks and growing the industry? Yes, especially via telematics, embedded and parametric insurance, on-demand models, micro-insurance, etc.). The net here is that the application of advanced technologies has allowed the industry to process over three times the premium since 2000 with similar numbers of producers and underwriters. (The number of P&C underwriters has grown from the mid-60,000s into the 70,000s, although exact numbers are elusive.)
Predictions for AI and P&C insurance in the next decade
Am I building a case that AI will not change P&C insurance much over the next decade? Absolutely not! As we know, the insurance industry has been leveraging AI technologies for quite some time, including machine learning for predictive models, computer vision for aerial imagery related to property risks, chatbots for customer services, and more. But today, AI is driving tremendous changes, particularly surrounding generative and agentic AI for a wide variety of use cases across the value chain and ecosystem. Large carriers are unleashing GenAI tools to tens of thousands of employees and seeing impressive productivity gains. Vendors are embedding AI into their solutions for agency management systems, CRM, policy and claim systems, and underwriting platforms. New insurtech players that are AI-native are bringing impressive capabilities to the industry.
Given these developments and this trajectory, what are we to expect from the industry in 2035? Here are some predictions with rationale:
Premium growth: Expect P&C premiums to double by 2035 to around $2T. This is slightly higher than other industry projections but still sits comfortably at about 7% compound annual growth rate. My prediction is not just based on rate increases but also on an expanding pie of risks. The industry will increasingly be able to address personal and business risks that have heretofore been unaddressed (essentially self-insured).
Shifts in employment and roles: My prediction is that key roles such as agents and underwriters will see a slight decline. At a time when the Bureau of Labor Statistics is predicting big job losses of 20-30% or more in other professions, their forecast is that there will only be 3-4% fewer underwriters by 2035. Underwriting assistants and processors may be a different story as this is where AI is likely to hit hardest. However, the biggest change will be the evolution of roles and the emergence of new roles. Underwriters will be more focused on relationships and managing the portfolio. Adjusters will spend more time on relationships and complex claims. Agents will migrate to more activity on the commercial
lines and specialty side, especially as risk continues to become more complex and nuanced. Ultimately, I believe it is possible to be both pro-AI and pro-human, with AI elevating the roles of industry professionals.
Greater efficiency and productivity: The net result of premium growth projections and limited decreases in agents, underwriters, and adjusters will be major increases in productivity. Handling two or three times (or more) as many customers/accounts/cases will be the norm.
Innovation: AI will significantly increase speed to market and innovation. Expect to see more variation in products and programs, more rapid in-market adjustments, and faster response to emerging risks.
Evolution for partners: Partners and providers to the P&C insurance industry will be significantly affected by AI, with major implications for Third Party Administrators and Business Process Outsourcing providers. Despite this, there will still be a strong need for these providers to deliver services to the industry, especially in light of the widening resource gap – as highlighted above. Plus, notwithstanding the impact of AI, estimates show that the P&C industry will have a shortage of at least 400,000 professionals by 2035. TPA and BPO firms with deep industry expertise will continue to be good partners to the industry. At the same time, these firms must tech-and AI-enable their offerings to the industry to continue to improve the effectiveness, productivity, and ROI of their services.
Conclusion
The fundamental business model of the P&C insurance industry is unlikely to change. The role of the industry in the world will be heightened, and its professionals will enjoy fulfilling careers. There will certainly be short-term pain with job loss of entry-level and more administrative type employees, but new roles will be created for those that are willing to learn and adapt.
There will clearly be winners and losers among companies and employees. But the overall picture for the industry will remain bright in the AI era.
Explore the full research behind these trends.
