Published
In a recent fireside chat, Genpact's insurance leaders Yasir Andrabi and Don Okolie sat down with HFS Research to discuss one of the insurance industry's most urgent problems: why decades of digitization have not materially improved underwriting outcomes, and how agentic AI is reshaping underwriting judgment, risk selection, pricing adequacy, and portfolio quality.
This article distills the most important insights from that dialogue, selected to reinforce the themes that underwriting leaders care most about today.
The reality check: Digitization didn't fix underwriting
Despite massive investments in core platforms, workflow engines, and analytics, underwriting outcomes remain stubbornly unchanged. Underwriters still spend nearly half their time on noncore work. Recent HFS research shows that nearly 90% of underwriters expect real-time, data-driven underwriting to replace traditional models within five years, yet fewer than a third are structurally ready to support this.
Modernizing underwriting workflows didn't improve outcomes because underwriting is not a linear sequence of tasks. It's an interplay of prioritization, context, trade-offs, risk interpretation, pricing discipline, and portfolio sensitivity.
The core insight here? Digitization improved processes – not judgment. And underwriting is, at its core, a judgment discipline.
Faster workflows did not – and cannot – solve the central problem: underwriters are making judgment calls with fragmented, inconsistent, incomplete information.
Agentic AI: Orchestrating decisions, not just automating steps
So what will separate those who use AI from those who change their economics? The leaders who materially impact underwriting profitability will:
Move AI upstream, letting it intervene before human judgment is applied
Shift to confidence-based decision models
Continuously evolve decision boundaries and feedback loops
Let AI reshape how underwriting works, not just how fast it runs
The winners will be those who architect underwriting as a self-improving decision engine. And that's where agentic AI fundamentally changes the game.
Unlike traditional automation or rules-driven AI, agentic AI intervenes where underwriting performance is determined: in the decisions themselves. From the conversation, three traits emerged as the defining markers of a true agentic underwriting system:
1. Intent-driven action, not predefined workflows
Agents determine what to do next when pathways are not preset.
2. Context accumulation and memory
The system learns from outcomes, broker behavior, geography, past claims, and emerging patterns.
3. Autonomous delegation of judgment
Within defined boundaries, agents decide when to act versus when to escalate to a human underwriter.
These characteristics enable AI not only to speed up work but also to supercharge underwriting decisions with real-time clarity, consistency, and context.
How Genpact Insurance Policy Suite fits – powerfully and by design
Underwriting transformation does not always require core replacement. Genpact Insurance Policy Suite sits as a modular, headless decision intelligence layer that enriches the existing ecosystem of core policy administration systems and applications.
Genpact Insurance Policy Suite supercharges underwriters via four modules, all powered by agentic AI, all API-driven, and all designed to scale without multiyear platform transformations:
1. Submission Clearance: Eliminating wasted human capacity
Agentic systems can analyze up to 100% of submissions, instantly identify appetite mismatches, and give brokers rapid science-backed feedback, freeing underwriters from the noise.
2. Risk Assessment: From intuition-based to evidence-based
Underwriters often rely on tacit knowledge and intuition when data is poor or incomplete. Agentic systems reverse this by:
Analyzing 500+ configurable risk factors
Benchmarking against your other portfolios
Detecting anomalies
Modeling loss behavior
What was once invisible becomes visible.
3. Exposure Management: Not slower – smarter
Exposure management shifts from being a modeling exercise to a real-time decision discipline. Underwriters get:
Near real-time exposure views
Accumulation flags
Tail risk indicators
Portfolio stress correlations
The result: decision-grade confidence at bind.
4. Quote & Bind: The "super underwriter" experience
Underwriters receive dynamic recommendations on:
Limits
Deductibles
Clauses
Appetite alignment
Real-time portfolio impact
This brings consistency, transparency, and portfolio discipline into every decision.
This article is based on a recent interview with HFS Research for their Unfiltered Stories series. Watch the full discussion here.
Want to stay ahead in insurance? Get insights and expert tips — straight to your inbox
