A first-of-its-kind RGA collaboration with AWS at an AHOU 2026 session revealed a clear message: The future of underwriting is already taking shape, with strong alignment around data, AI, and speed as the defining trend. What remains unclear is how organizations will manage the implementation and the implications, particularly around governance, accountability, and the role of human expertise.
The insurance industry is aligned – and uneasy.
At a recent Association of Home Office Underwriters (AHOU) session co-led by RGA entitled “Underwriting 2050: Shaping the Future Today,” a new approach to capturing industry insight was tested in real time for the first time. Rather than relying on traditional polling, participants in the session were invited to submit full-text responses to a series of forward-looking questions. These responses were captured using an AI-enabled platform developed through a collaboration between RGA and Amazon Web Services (AWS).
The innovative approach used at AHOU 2026 enabled unconstrained input, allowing underwriting professionals to express perspectives in their own words without predefined categories or limitations.
Responses were aggregated and analyzed in real time using AI to:
The results were unusually clear: The industry is no longer debating the direction of underwriting; instead, it is debating the consequences.
Across every question – from future drivers to long-term risk – the same themes repeatedly emerged:
Despite this alignment, exactly what happens next and what it will mean for the industry remains up for debate.
The strongest signal in the research was not new, but the level of consensus revealed further progress in underwriting’s ongoing evolution.
References to “data-driven underwriting,” “complete and accurate data sources,” and “integration of multiple data streams” dominated responses.
What has changed is how data is being framed. It is no longer described as a tool to improve underwriting but as the foundation of underwriting itself.
This distinction matters. When underwriting becomes data-centric:
At the same time, this shift introduces new forms of risk. Multiple responses pointed to concerns about “mortality slippage” and “anti-selection,” highlighting an uncomfortable truth: Better data does not guarantee better outcomes. Instead, it changes where the risk resides.
There is broad agreement that AI will transform underwriting. Respondents consistently referenced:
But beneath that agreement sits a clear tension. Some responses suggest human underwriters will largely disappear. Others explicitly argue the opposite.
This indicates a transition, not a contradiction. What the research makes clear is that the industry understands what AI can do. However, it does not yet know:
This gap between capability and operating model is where risk is accumulating.
Across responses, a consistent future emerges. Routine underwriting becomes automated, and human underwriters focus on exceptions. This creates a bifurcated model:
In this model, the role of the underwriter changes in three fundamental ways:
The result is a sharper role, not a smaller one, for underwriters, and that role is much harder to train.
When asked directly about risks, the tone of responses shifted: The concern was not that AI would fail but that it would succeed. Specifically:
This is not a theoretical concern. If underwriting becomes primarily automated:
One participant captured the tension clearly: “Having everything analyzed and summarized for an underwriter erodes the profession’s risk selection.”
This reveals a capability issue. The real risk is not job displacement; instead, it is the loss of the skills underwriting depends on.
Another signal cuts across every question: speed. Terms such as “real-time,” “instant decisions,” and “frictionless experience” appeared repeatedly, often in the same response as concerns about accuracy and control.
Historically, underwriting balanced two competing priorities:
That trade-off is no longer acceptable. The expectation now is both:
Building on that is another key: Decisions from AI systems must be explainable and defensible. This creates a new operational standard. It is no longer sufficient to optimize a single capability. Organizations must deliver speed, precision, and transparency simultaneously.
While the research reflects broad uncertainty, the implications for action are relatively clear. Leading organizations are not waiting for perfect clarity. They are building around what is already known.
Three shifts stand out:
These actions help redefine how underwriting works. They do not speak merely to adopting new technology.
The future of underwriting is no longer a matter of pure speculation. Across hundreds of real-time responses, the direction is clear: Data-centric, AI-enabled, and built for speed.
What is not clear, and what ultimately matters more, is how organizations choose to operate inside that future. The tension running through the research stems from what may be lost in the transition for underwriting as a profession:
Execution is where these questions are resolved, and it is where differentiation will emerge. The organizations that succeed will not only adopt new technologies; they will redefine underwriting as a system — integrating data, AI, governance, and human expertise into a model that is both fast and resilient.
The industry is aligned on where underwriting is going. The question is whether organizations are prepared to operate there.
Explore the full “Underwriting 2050” white paper for detailed insights, methodology, and complete findings on how underwriting is being reshaped at scale.
