ITC Vegas 2026: How incumbents are restructuring to run on AI
2 October, 2026
Building AI-Ready Organisations
At the ITC Vegas Conference 2026, Partner Manmeet Singh Bawa hosted the Day One Panel ‘Rewiring the org: How incumbents are restructuring to run on AI’.
Read his key takeaways from the panel discussion:
AI remains a major focus at ITC Vegas 2026, but the conversation has shifted from excitement about the technology itself to a harder question: where is the demonstrable value, and how can insurers move beyond pilots to sustainable, repeatable deployment? The panel identified three critical elements of the “north star” for transformed carriers: people, processes, and governance. These need to evolve together; deploying new technology without adapting the surrounding operating model is unlikely to produce lasting value.
Talent requirements will materially shift in the coming years. One panelist noted that candidates are now asked during interviews how they would use AI to perform parts of their roles – an indication that AI fluency is starting to become an expected workplace capability rather than a specialist skill. Data literacy will similarly become a core capability alongside essential business skills such as underwriting. We therefore expect underwriters to become more data versed and increasingly comfortable using AI-supported insights while retaining the judgement, domain expertise, and relationship skills that sit at the heart of the role. The emerging model is not one in which data analysts simply replace underwriters, but one in which underwriting expertise is augmented by stronger data and technology capabilities.
The value of AI remains an open question. One panelist described a pilot whose technology environment and assumptions differed structurally from production, undermining the pilot parameters and creating issues when the solution moved toward launch. The example reinforced an important lesson: proving that technology can work in a controlled pilot is different from proving that it can work reliably within the realities of an insurance operating environment. Oxbow Partners’ view is that successful carriers will therefore run pilots against clear business outcomes rather than experiment with technology for its own sake. They will also build strong feedback loops around those pilots – capturing what worked, what did not, and why – so that unsuccessful experiments create useful organizational learning and reduce unknown unknowns before scaling. The objective should be to move from disconnected proofs of concept to repeatable deployment that improves end-to-end workflows and creates compounding value over time.
AI governance is often seen as slowing progress, but the panel agreed that it is essential given the nature of the industry. The challenge is to design governance that provides appropriate oversight without imposing the same burden on every initiative. Leading practice therefore points to “proportionate governance”: everyday AI use cases and embedded or higher-impact AI use cases should be evaluated differently, with governance calibrated to the significance, complexity, and potential risk of the application. This allows organizations to encourage experimentation where the risk is low while applying greater discipline to initiatives that affect core workflows, material decisions, or broader enterprise capabilities.
On operating models, panelists described several approaches. One is a central “air traffic control” model that establishes enterprise guidelines, guardrails, and coordination while enabling business units to develop their own data and AI strategies. Another is a central “delivery” model in which the enterprise exercises greater control over AI implementation rather than leaving delivery to individual business units. The right balance will vary by organization, but both models highlight the importance of clear ownership, decision rights, and coordination between business and technology teams. All panelists agreed on one point: business priorities should drive use-case selection and prioritization, with technology serving as a critical partner in delivering outcomes. The practical implication is that AI strategy cannot sit apart from business strategy; it needs to be anchored in the problems carriers are trying to solve, the capabilities they want to build, and the measurable outcomes they expect to achieve.
About the author
Manmeet Singh Bawa is a Partner in the US Office, leading the firm’s US business from New York. He brings over a decade of global consulting experience, advising leading insurers and multinational organisations on strategy, innovation, and business model transformation.
Manmeet has deep expertise across the P&C insurance sector, supporting clients on growth strategy, business model innovation, operating model transformation, embedded insurance, ecosystem partnerships, and data-driven performance improvement.
A recognised thought leader in the industry, Manmeet has published extensively on insurance innovation and is a regular speaker at industry events.
