Wins, Pitfalls, and Best Practices in Integrating AI in Carrier Health Underwriting: Insights from Industry Leaders
As AI becomes increasingly embedded in healthcare, group health insurance carriers are looking to leverage it to enhance underwriting, optimize risk assessment, and improve portfolio performance.
Gradient AI recently sponsored a webinar, The Role of AI in Shaping Carrier Health Underwriting: Proven Approaches and Insights hosted by Becker’s Payer Issues, in which insurance leaders shared their journeys, lessons, and outcomes around integrating AI into their underwriting process.
>> Watch the on-demand webinar now
Expert Panelists Share Their AI Integration Experiences
Our expert panelists from the group health insurance industry:
- Matthew Agdish, AVP Underwriting, Blue Choice Health Plan
- Nancy Schiel, VP of Underwriting, Commercial Markets Division, Medica
- Josh Phelps, Solutions Consulting Group Health, Gradient AI
Matthew Agdish has over 30 years industry experience with National and Regional Carriers and underwritten groups of all sizes and funding options. Nancy Schiel also has extensive experience, with over 30 years in the business. She is currently responsible for developing underwriting and pricing strategies for single employer large groups, fully insured and self-funded products.
Prior to joining Gradient AI, Josh Phelps served in multiple health insurance roles, most recently as Chief Actuary for a major blues plan. Josh’s area of expertise includes actuarial pricing, large and small group underwriting, machine learning, programming and artificial intelligence.
7 Key Insights About AI Adoption: Learn from the Webinar Panelists
This blog post summarizes 7 key insights from the webinar, along with direct quotes from our panelists, plus their actionable takeaways.
1.
AI is Now a Must-Have for Carriers:
From staying competitive to refining pricing strategies, insurers realize that AI is no longer optional – it’s a strategic imperative.
Matthew Agdish: “When I came into this leadership role, we didn’t have any predictive analytics. That put us behind. Now with Gradient AI’s predictive analytics solution, we’re actually ahead of the curve.”
Key Takeaways:
- AI helps catch up or leap ahead in competitive markets.
- Predictive analytics is now essential for assessing risk more accurately.
- Carriers not using AI are already seeing the consequences of being behind the curve.
2. AI Provides Better Risk Identification and Portfolio Management:
AI tools allow insurers to go beyond surface-level demographics to uncover deep, actionable insights about group health risks.
Nancy Schiel: “We had a tight corridor of average groups. AI helped us finally identify good-risk groups more accurately, which was the missing piece.”
Key Takeaways:
- AI enables more granular risk segmentation, not just high-risk avoidance.
- Underwriting becomes more consistent, with reduced person-to-person variability.
- First-year renewals are now more predictable and stable.
3. Common Pitfalls and Lessons Learned About AI Integration:
Adopting AI comes with challenges, including these three: over-reliance on risk scores without knowing what’s driving the results, poor calibration with your cost structure and book of business, and slow implementation.
Critical insights from the panelists:
Josh Phelps: “Slow adoption has been a major misstep in the market. Some have waited too long, and now they’re trying to catch up.”
Matthew Agdish: “One of our early mistakes was not benchmarking AI outputs properly. If you put too much or too little weight on it, you risk mispricing.”,
Nancy Schiel: “It’s about guardrails. You need boundaries around the tool, so underwriters don’t go rogue.”
Key Takeaways:
- Don’t blindly follow AI. Instead, balance it with actuarial judgment.
- Calibrate models to your own cost structure and data reality.
- Delays in AI adoption can leave carriers exposed or competitively disadvantaged.
4. AI is Not a Replacement for Seasoned Underwriters:
A common concern is that AI may replace human jobs. Our panelists told us that in reality, it’s more about augmentation than automation or replacement.
Nancy Schiel: “There was initial concern about job security. But our team quickly saw how AI improved their effectiveness.”
Matthew Agdish: “The consistency AI brought took out a lot of guesswork. It helped our underwriters make better decisions without feeling replaced.”
Key Takeaways:
- AI reduces subjectivity and variability, improving confidence in decisions.
- Underwriters feel more empowered when using AI tools, not threatened.
- It frees time for value-added analysis, rather than repetitive tasks.
5.
Data Scale and the “Build vs. Buy” Debate:
Some carriers considered developing their own in-house AI tools. But the depth and breadth of vendor datasets tip the scale toward buying. Vendors with access to massive datasets (i.e., millions of members) can train far more effective models than most in-house carrier systems. This makes buying AI tools more practical and powerful than building from scratch for many insurers.
Josh Phelps: “Carriers might have tens of thousands of claims. At Gradient AI, we have over 15 million lives in our models. That level of scale matters.”
Key Takeaways:
- External tools often offer more robust models, thanks to larger datasets and refined algorithms.
- The build vs. buy decision should account for speed to market and data breadth.
6.
AI Helps with Funding Strategy Optimization:
AI doesn’t just help with rating. It also informs funding mechanisms, helping determine whether a group is better suited for fully insured, level-funded, or self-funded plans.
Josh Phelps: “AI helps identify the right fit for a group’s risk profile, which leads to better long-term outcomes for both the client and the insurer.”
Key Takeaways:
- Smarter funding recommendations = better client retention and more sustainable books.
- Helps brokers and clients understand the “why” behind the rate without exposing raw outputs.
7.
Practical Advice for New Adopters Getting Started with AI:
The panelists had plenty of guidance for carriers just beginning their AI journey.
Nancy Schiel: “Do your own analysis and understand the data. Partner with your actuarial team and make sure you have calibrated appropriately based on your cost space and your book of business.”
Josh Phelps: “Focus on results-driven outcomes. Trust and verify. Have your vendor do a Proof of Concept or Pilot so you can see how this solution would have affected your books historically or how you would use it going forward and compare the impact if you did not have the tool.”
Matthew Agdish: “Make sure the tool you use has the ability to drill down and provide you with additional details. It’s nice to have the scores and the dashboard of results on a high-level but having the statistical detail we get with Gradient AI, that level of insight and detail is invaluable.”
Key Takeaways:
- Develop internal AI fluency: understanding how to use AI matters just as much as having it.
- Run a proof of concept or pilot before full rollout.
- Do your homework. Focus on outcomes, not just hype. Make sure your solution gives you the data behind the risk score.
- Read Gradient AI’s detailed Guide to Change Management around AI adoption, for a clear roadmap of actionable steps for successful integration across your teams.
The Most Successful Carriers are Those Who Combine AI with Sound Underwriting Judgement, Good Data Practices, and Organizational Alignment
Whether you're at the start of your AI journey or refining an existing approach, one message is clear: AI is here to stay, and those who embrace it carefully and smartly are reaping the rewards.
To Learn More, Watch the On-Demand Webinar!
To hear more from the panelists, watch The Role of AI in Shaping Carrier Health Underwriting: Proven Approaches and Insights, hosted by Becker’s Payer Issues, and sponsored by Gradient AI. To learn more about AI for Carrier underwriting, contact us today.
Becker’s Payer Issues is the leading news organization covering the U.S. health insurance industry. The editorial team delivers daily breaking news, business analysis and expert interviews across the evolving insurance landscape, delivering the latest need-to-know.
Stay on top of AI trends by subscribing to Advanced Insights, the newsletter for strategies, ideas, and insights on AI in insurance. Delivered monthly to your inbox. Subscribe Now →