Speaking & Media

For most U.S. banks, AI is no longer a question of pilots and proofs of concept — it's a question of how to turn isolated experiments into enterprise capability that moves the financials. This session maps the path from AI POC to enterprise AI: what has to change in the operating model, data foundation, and governance when intelligence moves from the edges of the organization into the core of banking, payments, risk, and customer engagement.

The discussion is organized around four shifts that separate leaders from the rest. First, moving from AI POC to enterprise AI — productionizing and scaling what already works. Second, moving beyond productivity to enterprise value — tying AI to revenue, margin, and return rather than activity. Third, secure and responsible AI — deploying in ways that hold up to regulatory, risk, and model-governance scrutiny. And fourth, tokenomics and AI financial optimization — managing the economics of compute so AI spend is metered, concentrated on the highest-leverage use cases, and run like a disciplined investment.

In the span of twelve months, AI-generated code went from a minority practice to a majority one, and autonomous agents began shipping production work at scale — even as token costs rose 5–9x in a single model generation and compute, on average, began to cost more than the teams it was meant to augment. The result is a widening gap between two compounding curves: a productivity curve that rewards adoption, and a cost curve that punishes undisciplined spend. The winners are no longer the firms with the best models; they are the firms with the operating discipline to capture the productivity curve while staying ahead of the cost curve.

This session reframes AI value capture as a CRO and COO scoreboard rather than a technology decision. Working through real adoption and cost data, it shows why AI is uneconomic at the average yet highly profitable at the top of the leverage curve — and how disciplined firms concentrate budget on their highest-leverage talent and use cases, meter consumption like a budget rather than electricity, and continually renegotiate the build-versus-buy line as economics shift. Executives leave with a practical framework for positioning their organization on both curves, and a structured way to move AI out of technology and into revenue, operating margin, and return on invested capital.