Insurance Leaders Put the Difficult Questions About Agentic AI on the Table

Artificial intelligence in insurance is moving from experimental projects into the systems, controls and operating decisions that shape day-to-day business. That shift is bringing a more demanding set of questions to the surface: who governs autonomous agents, how should their decisions be tested, and what does responsible deployment look like at scale?

Those questions will be at the centre of Agentic and Generative AI for Insurance USA, a two-day conference taking place at The Westin New York at Times Square on 18–19 November 2026.

Organised by Intelligent Insurer, the event is expected to bring together senior insurance executives, technology specialists, transformation leaders and AI practitioners. Alongside its keynotes and case studies, ten roundtable groups will give delegates an opportunity to examine the operational challenges behind the industry’s AI ambitions.

The conversation has moved beyond experimentation

The central premise of the programme is that insurers are no longer asking only what generative and agentic AI can do. Attention is turning towards the infrastructure, governance and organisational design required to use it safely and consistently.

Figures cited by the organisers from Evident’s Q4 2025 Insurance AI Use Case Tracker suggest that 68% of publicly disclosed insurance AI deployments were generative or agentic. Agentic systems alone represented 21% of disclosed deployments.

That does not mean the difficult work is complete. A successful pilot can demonstrate technical potential without answering questions about accountability, model drift, data access, customer trust or the consequences of an agent acting incorrectly.

“The biggest risk is architectural debt disguised as innovation velocity.”

Clayton D’Souza, Vice President of IT Architecture at Old Republic Commercial Risk, will host a discussion on AI governance in practice. He argues that insurers need the appropriate controls—including identity, auditability, explainability and mechanisms for stopping autonomous actions—before they attempt to scale agentic systems across an enterprise.

Ten questions insurers are now confronting

The roundtables will explore a range of decisions that are becoming increasingly important for carriers and their technology partners:

• Whether to build, buy or configure agentic AI systems

• How AI agents should be monitored and reinforced

• What effective governance looks like in practice

• How systems should be tested and validated

• How organisations can manage model drift and explainability

• Whether AI will disintermediate or transform the broker relationship

• How successful pilots can be scaled across an enterprise

• What customers need before they will trust AI-enabled services

• How agentic systems could change portfolio management

• How workforce insight can support adoption rather than resistance

These are not purely technical questions. Each one affects procurement, organisational design, customer experience, risk management and the evidence leaders will need when deciding whether an AI investment is delivering measurable value.

Rethinking the business around AI

Joe Barneson, Head of Engineering at MGT Insurance, will lead a roundtable examining the future of the broker channel. His view is concise: “It’s transformation, full stop.”

The distinction matters. Adding an AI tool to an established workflow may improve the speed of an individual task, but it does not necessarily change the economics or structure of the business. The larger opportunity—and the greater challenge—is to reconsider processes around what the technology now makes possible.

For established insurers, that can mean adapting businesses designed for a very different operating environment. Newer organisations may have more freedom to design around AI from the outset, but they still face the same questions about trust, control and sustainable value.

A format designed for candid discussion

The roundtables will run across two rotations on the first day of the event. Hosts include practitioners from insurance, investment, data science, engineering and technology roles, with the discussions intended to offer a more candid setting than a conventional stage presentation.

Helen Raff, Director of Events at Intelligent Insurer, said the format was designed to let participants discuss what is not working as openly as what is. That includes situations in which governance has fallen behind adoption, build-versus-buy decisions have produced unexpected consequences, or AI has begun to test existing broker relationships.

More than 200 senior AI specialists and over 30 C-level speakers are expected to participate across the wider programme. Sessions will focus on practical deployment, responsible scaling, operating models and measurable returns.

Event details

Agentic and Generative AI for Insurance USA takes place at The Westin New York at Times Square on 18–19 November 2026. Further details and registration information are available from the event organiser: https://events.newton.media/Agentic-and-Generative-AI-for-Insurance-USA

This article is based on information supplied by Newton Media through Pressat. It has been edited by Expert View Media for clarity, context and length.