How are you approaching generative AI adoption in your organizations? Are you focusing first on quick-win use cases (like chatbots/automation) or deeper integrations into core platforms?

Great question. I've taken a quick-win approach to introducing AI, but the caveat here is that I've taken that quick-win step directly with the business. I chose two specific business partners (i.e., Finance and Legal) and asked them a simple question: "What could my team do to help you make quicker and more informed decisions as part of your day-to-day deliverables?" The answer was as simple as the question: "[From Finance]... I need a quick way to discover, consolidate, analyze and present data (i.e., find/search, data/reports and outputs) against a series of complex scenarios"; "[From Legal]... I need a summary of all change of control provisions found in Legal documents (i.e., any restrictions on transfer of shares, assets or operations), and the like". So, I embarked on a discovery of my own to determine if I could find an all-in-one AI solution provider who would be prepared to offer a free trial/pilot of their service to dip our toes in the waters of AI. We are now embarking on a control experiment (with all the necessary security protocols, compliance, and governance measures in place) to have the business assess the AI capability that IT recommended. We've just started, so there's not much to share yet. Still, the point of the exercise is to create a buzz around the business, forming acceptance, endorsement, and sponsorship of an IT-recommended solution [AI, in this case] based on having forged a strong partnership between IT and our business stakeholders. This way, the business gets behind the innovation, not just IT. Plus, IT has a better chance of securing a budget to fund the IT investment required to develop, test and implement it.

Internal use cases should be handled differently from external-facing solutions. While they may not require the same level of polish, they still demand rigorous testing and man-in-the-middle validation to ensure accuracy. Solutions like GenAI-powered knowledge bases and certain automations are relatively easy to build, but all implementations must adhere to compliance, security, and data governance standards.
We've achieved several quick wins using GenAI, and a few initiatives have progressed into more complex integrations.