But when you guys looked to start this whole process, what were the sort of end goals for

It's really interesting when I think back to when we started because it was so new and there was so much unknowns and there wasn't really a lot of case studies we could learn off or there wasn't a lot of resources out there.

So initially we just, as soon as what was called Bing Chat Enterprise at the time, now Copilot came out, I think within a few weeks we rolled that out across the firm.

And our goal at that time was quite broad, just to educate people on AI.

And this is something we've had to constantly evolve and pivot from because... as the tech evolves, as the needs evolve, as the regulatory landscape evolves, we have had to kind of adjust the way we work.

So we really just wanted to have that general capability uplift.

And looking back, we didn't identify this at the time, but now we can see this is what we were trying to do is really firm up that baseline AI literacy across the firm, across all the teams, across all of the role levels.

So just to clarify on that point, it sounds as though instead of it being a top-down approach where you're dictating use of these new technologies into daily workflows and operations, it was more about trying to hear the feedback from the ground up so that as a company, you could then make broader and more holistic decisions about what was going to constitute best practice.