Roundtables put insurers' toughest AI questions on the table at Agentic and Generative AI for Insurance USA

Taking place at The Westin New York at Times Square, the two-day event will explore how insurers can move beyond promising pilots and embed agentic and generative AI into the enterprise infrastructure, operating models and workflows needed to deliver measurable business value.

Ten dedicated roundtable discussion groups, running across two rotations on Day One, will bring together architecture, data science, engineering, technology and innovation leaders to debate the questions carriers are wrestling with right now: build vs. buy vs. configure, monitoring and reinforcing AI agents, governance in practice, testing and validation, model drift and explainability, the future of the broker relationship, scaling from pilot to enterprise, customer trust, portfolio management, and turning workforce insights into AI adoption.

The roundtables sit alongside a wider program of keynotes, panels and case studies, and reflect a broader shift in how the industry is approaching AI. According to Evident's Q4 2025 Insurance AI Use Case Tracker, 68% of publicly disclosed insurance AI deployments were generative or agentic, with agentic AI accounting for 21% of deployments. The direction of travel is clear: AI is no longer experimental in insurance; it is increasingly operational — and the questions insurers are asking have shifted from what AI can do, to how to govern, scale and trust it.

Clayton D'Souza, Vice President, IT Architecture at Old Republic Commercial Risk, is hosting the “AI Governance in Practice” roundtable. He said:

“The biggest risk is architectural debt disguised as innovation velocity: organizations bolt autonomous agents onto legacy platforms and governance models built for deterministic systems, and the gap between agent capability and enterprise control maturity becomes the attack surface — for compliance failure, model drift, and reputational risk alike.

“The strategic imperative for architecture leaders is building the control plane first — identity, auditability, explainability, and kill-switches for autonomous action — so that scaling agentic AI is a governed capability rollout, not a series of ungoverned point solutions.”

– Clayton D'Souza, Vice President, IT Architecture, Old Republic Commercial Risk

On why he chose to focus his roundtable on enterprise readiness rather than the pace of AI adoption, D'Souza added:

“I want this industry conversation to move past ‘what can agentic AI do’ to ‘what does our architecture need to look like for agentic AI to be trustworthy at scale’ — that's the harder, more consequential question.”

– Clayton D'Souza, Vice President, IT Architecture, Old Republic Commercial Risk

Joe Barneson, Head of Engineering at MGT Insurance, is hosting the “AI + Brokers: Disintermediation or Transformation?” roundtable, which will debate what agentic AI means for the broker channel. He said:

“I believe it's not disintermediation vs. transformation; it's transformation, full stop. Insurance is particularly exposed because AI has the potential to change not just individual tasks but the underlying way companies operate.

“For incumbents, that means figuring out how to reshape businesses that were built for a very different operating model, while newer companies can design around AI from the start. The winners in this next wave will be the ones who adopt AI and build around it, not the ones who fear it or try to box it in.”

– Joe Barneson, Head of Engineering, MGT Insurance

On the pace of adoption, Barneson added:

“The pace is uneven, and I think many companies still underestimate how quickly the technology has moved. I recently had a vendor tell me a change would take months, and within three days I had used agentic development to augment their platform to do what I needed myself. Most of the industry is still thinking about how to bolt AI onto existing processes, when the bigger opportunity is to rethink how the business operates around what is now possible.”

– Joe Barneson, Head of Engineering, MGT Insurance

The roundtable format is designed to give delegates a candid, off-the-record space to test ideas with peers facing the same operational and governance pressures, rather than watching a stage from a distance. Hosts leading the ten sessions include Ramesh Sethi (Prudential Financial), Ravinath Kausik (The Hartford), Michael Mocanu (Liberty Mutual Investments), Ghanshyam Upadhyay (Gallagher Bassett), Rayn Cole (formerly Everest), Darrel Barbato (The Hanover), Crystal McQueen (Great American Insurance Group) and Lisa Hollingshed (Nationwide), alongside D'Souza and Barneson.

Helen Raff, Director of Events, Intelligent Insurer, said:

“The roundtables are where the real conversations happen. On the main stage, leaders share what's working. Around the table, they're able to be far more candid about what isn't — where governance is still catching up, where a build-vs-buy decision went the wrong way, where a broker relationship is being tested by AI faster than anyone expected.

“We've deliberately built distinct roundtables, hosted by practitioners rather than vendors, because the challenges in scaling agentic AI look different depending on whether you sit in architecture, data science, engineering or innovation. Bringing those perspectives into the same room, in a format built for honest exchange rather than polished soundbites, is exactly what this industry needs right now.”

– Helen Raff, Director of Events, Intelligent Insurer

With more than 200 senior AI experts expected to attend and over 30 C-level speakers on the agenda, Agentic and Generative AI for Insurance USA is designed for senior decision-makers focused on real-world deployment, scalable use cases and measurable ROI. Alongside the roundtables, the program features keynotes and panels from speakers including Helene Kubon Skulstad (Liberty Mutual Insurance), Jay Bell (Zurich North America), Kyle Ramsay (Hippo Insurance), Leyla Delic (AXA XL), Andrew Leeds (Plymouth Rock Assurance), Prashant Hinge and Ronald Morrison (MSIG USA), and Samrat Dua (Swiss Re).

With an agenda focused on practical deployment, responsible scaling and measurable outcomes, Agentic and Generative AI for Insurance USA is an essential event for insurance leaders seeking to build the capabilities, operating models and competitive advantage required for the next phase of AI transformation.

For more information and to secure your place, visit Agentic and Generative AI for Insurance USA: https://events.newton.media/Agentic-and-Generative-AI-for-Insurance-USA 

Did you get value from this story? Sign up to our free daily newsletters and get stories like this sent straight to your inbox.