Over the last couple of posts, we walked through what the EDM AI Change Assistant is, how it can be helpful, and seen some of what it can do so far. If experience has taught us anything, it’s that Oracle will continue to develop the tool and add additional features over time. I think the important thing to note is that the value of governance hasn’t decreased because we have AI to help out. If anything, I think it makes governance more important.

The assistant makes it easier to find your nodes, build queries, and create request items using prompts. Some of the careful clicking and dragging will hopefully be a thing of the past which is a fantastic improvement for admins and users.

But, as we saw in the last post, the AI Change Assistant doesn’t take credit or assume responsibility for the change, that’s on you. It’s not about whether the assistant can build a request, the question is whether that request is correct and/or should be submitted at all. That’s where having a human-in-the-loop is essential.

Back to governance for a second, if I can ask the assistant to make a bunch of changes and not review each one carefully that’s an open window for garbage to enter your system. If anything, I’ve learned that it’s going to be more important for us to start adding in additional custom validations to ensure that the business logic that lives in our heads is actually present in the system. This is the only way that we can flag when a user has given AI an incorrect prompt and violated a numbering sequence or naming convention.

While EDM contains data (metadata), it doesn’t know what humans do about the organizational politics or any future reorganization plans. Human reviewers bridge that gap. The last thing we want is AI to blindly accept what a user is asking it to do and/or humans allowing the system to arbitrarily make those changes without review.

Human-in-the-loop is a feature, not a bug. It’s exactly how the system should work to maintain integrity especially with your organization’s master data.

The Change Management Assistant is a significant step forward toward allowing administrators to spend less time constructing requests and more time evaluating their impacts. EDM has always been about managing change in a controlled manner. The arrival of AI doesn’t change that, it simply makes the need for human oversight and governance controls built into the system even more apparent.