Securities Regulation

The operation.

A national securities and capital-markets regulator in West Africa, partway through a leadership programme run with an external academy. The room: the acting head of IT and four divisional heads, covering the business office, service management, IT security, and enterprise architecture.

Senior, technically literate, and two days into strategy slides. They wanted something live and specific to their own perimeter.

The brief.

The leaders wanted two things. First, to understand how to use AI practically, as a productivity tool in their own working day. Second, to see it applied to work they already had in progress: reviewing policy, checking compliance, and developing their digital-asset framework.

Two constraints shaped the day. Anything generic would be rejected in the room. And the engagement letter barred any internal regulator data from going into an AI tool, so every demo had to run on public documents.

What we did.

We loaded the public source documents live, in front of the room: FCA and MAS guidance on AI, and the regulator's own published digital-asset framework. The AI read them, compared them, and cited where each point came from. When a delegate pushed back on a detail, we opened the source together.

Four demos in the morning, each teaching one capability: a multi-source research brief on how peer regulators approach AI, a side-by-side policy comparison, a brainstorm on the regulator's own framework, and a sticky-note race between the room and the AI. The afternoon turned it into action: a planning canvas, pressure-tested around the table, and 30-day follow-up emails drafted straight from it.

What the room walked away with.

• Four live demos of AI doing real regulator work, on public documents only.
• Zero internal data entered into any AI tool, in line with the engagement letter.
• 90-day commitments for each delegate, every action with an owner.
• Follow-up emails drafted from the action canvas, so the work didn't end with the day.