The gap regulators are already looking at

OSFI's E-23 and the FCA's AI principles share a common assumption: that the people overseeing model outputs can actually interrogate them. Not audit the process. Not sign the attestation. Interrogate the output itself, with enough understanding to push back.

That is a specific cognitive skill. It requires knowing how models fail, what they optimize for, and when their confidence is statistically meaningless. Most oversight frameworks assume this skill is present. Most workforce development plans have not addressed it.

The gap between what regulators expect and what oversight layers can actually do is real, and it is growing as AI is deployed faster than the people supervising it are trained.

What most AI training in financial services actually covers

The majority of AI training in banks, insurers, and asset managers teaches people how to use tools, how to prompt, how to integrate outputs into workflows, how to work faster. That is useful, but it is not what the regulations are asking for.

Using a model and being able to challenge it are different skills. One requires familiarity. The other requires a working understanding of where models go wrong, how to spot when a confident output should not be trusted, and how to make a judgement call when the model and your own read of the situation diverge.

Very few organizations are deliberately building that second set of skills. The training programs do not cover it, and the performance metrics do not reward it.

What Steve covers with this audience

Steve speaks to risk, compliance, L&D, and senior leadership audiences on the specific thinking skills that AI oversight requires, and why standard AI adoption training does not build them. The talk is grounded in how models actually behave, not in general caution about technology.

He covers the cognitive patterns that make people poor at challenging model outputs, the organizational conditions that make rubber-stamping likely even when guidelines say otherwise, and what deliberate development of genuine oversight capability looks like in practice.

The session is available as a keynote or a longer workshop format for teams where the regulatory exposure is direct.

Topics for Financial Services audiences

Steve speaks to financial services organizations on the following topics. Each can be delivered as a keynote, half-day workshop, or executive briefing.

Who books Steve

L&D directors, CHROs, risk and compliance leaders, CDOs at banks, insurers, and asset managers.

If you are planning an event and want to discuss whether Steve's work is a good fit, the fastest route is a short conversation. No pitch deck required.