Authority Migration: The Shift Leaders Aren't Tracking

Something quiet is happening in organizations deploying AI at scale — and most leaders are completely missing it. Authority is migrating. Not disappearing. Moving. From humans to systems, from managers to models, from individuals to interfaces. MIT Sloan's 2025 research calls it "intelligent choice architecture" — AI that doesn't just inform decisions, but reshapes who actually holds decision power. In this episode, Dr. Marilyn Carroll maps the three pathways through which authority migrates, why it stays invisible until something goes wrong, and how to audit your organization before the shift becomes irreversible. If you think you're still in charge — this episode will make you check.

[ Hashtags ]

#AILeadership #DecisionRights #AuthorityMigration #FutureOfWork #AIStrategy #ExecutiveLeadership #TwoDoctorsAndATwist #AIGovernance #OrganizationalDesign #LeadershipDevelopment #DrMarilynCarroll #AIAccountability #WorkplaceTransformation

Stay connected with 2 Doctors & A Twist – Just What the Doctor Ordered!

• Subscribe on Apple Podcasts, Spotify, or wherever you listen.
• Follow us on LinkedIn for clips, insights, and upcoming episodes.

Transcript

SPEAKER_00 0:01

Hello everyone and welcome back to Two Doctors in a Twist. This is our limited series. And as I stated, on each episode, Dr. Jamie is out doing this particular filming of this series on a different assignment. So we are going to continue on with our limited series. And today, on this episode, rather, we're going to talk about this is um episode three. And today we're going to talk about authority migration, the shift leaders aren't tracking. Yeah. Authority migration, the shift that leaders aren't tracking. So let me start with a simple question for you. Are you making the decision or your system or is the system making it for you? That's important. I've used AI now uh for the last almost three years. I've been using AI. And yes, I use it to streamline process, help me with streamlining processes, help me with uh research. It helps me with a lot of different things, but it never has the final decision in anything I do. I have that final decision. As a matter of fact, the way I use AI, I build my research materials first, I pulled those in. I don't just go to AI and ask it a question. And my systems I built uh and testing this and how I got to this point is uh an understanding that I needed governance in the process on any system I put together for any client I was billing it for. So this is not in theory. I wanted you to know that that this is actually in practice. Because right now in organizations across every industry, leaders are still, or they leaders really still believe they hold decision authority. Did you hear me? Did you hear me? But what they're not seeing, guys, is how much that authority has already shifted. So stay me with me for a moment. What I want to do is give you some research background. Because something quiet is happening in organizations deploying AI at scale. And most leaders, like I said earlier, are completely missing it. Authority is migrating, not disappearing, is moving from humans to systems, from managers to models, from individuals to interfaces. How so? Here's some research for you. MIT Sloan research calls it intelligent choice architecture. AI that doesn't just inform decisions, but reshapes who calls it intelligent choice architecture. Okay, that's what it's called. And AI really, really actually holds the decision. It doesn't actually hold the decision power. So there are three pathways through which authority migrates. And here's why it stays invisible until it goes wrong, and how really to audit your organization before the shift becomes in in um irreversible. Okay, it's not root, you can't reverse it. If you think you're still in charge, this episode will make you check that, okay? I hope it does make you check that, okay? Because here you go. Because right now, organizations across every industry, as I stated earlier, they still believe they hold decision authority, but they're not seeing what they're not seeing is how much authority has already shifted, not disappeared, not been uh not replaced, migrated. Okay. What do you mean, Marilyn? What are you talking about? Migrating. Well, as most leaders have no idea this happening, let me give you something concrete. Okay. We go back to the hiring manager. Yeah, Marilyn, you're back to the hiring manager. Um, a hiring manager sits down to review candidates. The AI system has already ranked them. The top five are presented. Yeah, this is happening right now. The manager reviews them, chooses one. Now, here's the reality. For the last six months, that manager has selected from the AI shortlist 97% of the time. So let me ask you again, who is actually making the decision? Because the manager feels they are. But the system is shaping the choice before it ever, before the manager ever sees it. Yeah. And that's where this gets interesting. How so? Well, when you're defining, as we define the shift, this is what we call authority migration. Authority migration is not automation, automation replaces task, okay? Right? Authority migration reshapes who holds influence over the decision. Yeah. Who holds the influence over the decision, over the decisions that are being made? And that distinction matters because in most organizations, leaders are still looking for replacement while the real shift is happening in influence. It's happening in influence. So again, I said mentioned earlier, uh research from MIT Sloan Management Review describes this as intelligent choice architecture. So systems that don't just support decisions, they structure how the decisions are made. And once that happens, authority begins to move. Yeah. So look at this. I want you to look at this, um, look at the data. So research shows that 70% of organizations now have AI and production and environments, but fewer than 10% have mature governance structures. Now, you may be saying, Marilyn, so what? So what? Well, that gap matters because it tells us something critical. AI is influencing decisions faster than organizations are designing structures to govern it. And when that happens, authority doesn't disappear, it shifts quietly. And this is exactly what I describe in human government AI. Again, this is a book that's coming out in April, April 15th, I think it is. And when systems began shaping the information environment, they began shaping decisions and authority flows. Authority follows. One, it happens gradually, no single moment is this occurring. It's gradual, guys. It's gradual. No announcement, no visible handoff. Oh my gosh. Um, just small shifts over time. So until suddenly the system is doing more than supporting, it's shaping. Now, number two, the outcomes still look fine, decisions still work, but results still come in. Nothing appears broken, so no one questions the process. But effectiveness can hide misalignment. Yeah. Something can be so effective in the short term, but there's a misalignment. Um, being a researcher, I look at at heart, I really am. I find myself researching everything. Just this morning, I researched what was happening with me and my body and my weight loss and things of that nature. And because I had been going through this for about a year, a little over a year, and my doctors were asking me questions, and yes, I was taking the GLP ones, and I was like, Well, what is happening? Everybody around me that was taking it were doing great, having a great success. But I was not, and I was not because of something in my body that was had happened years ago during childbirthing years, and so that in and of itself had an effect on how that GOP1 was affecting me. So no one was really mapping, and when I talked to the my physicians about it, they were like, oh, dumbfounded, what had happened? What was going on? So I took the time to research it myself to watch my own patterns and what was happening, and finally we came to it. And it was a simple fix. But not only that, not that it was affecting the weight loss, it was affecting something else. But my insurance at the time was paying for the weight loss when the GLP was actually helping with something else that women struggle with a lot. But more to come on that, I sent that to um, I'm sending a paper to uh the drug manufacturer of the GLP one I was taking, as well as the two doctors who had been treating me for my weight loss condition. So no one is mapping decision authority. Let's get back on track here, Marilyn. Org charts, reporting lines, workflows. Uh, but almost no one maps where decision influence actually lives. So authority moves without being tracked. Now, the three pathways of migration include authority doesn't move randomly, it follows patterns. Remember that it doesn't move randomly, it follows patterns. So I want you to take this down. Number one, recommending recommendation lock-in. Leaders begin to trust the system's outputs over time. They stop, they stop questioning them. Then number two, the process starts embedding, it becomes embedded in the day-to-day workflows. AI becomes part of the workflow. You can't make the decision without going through the system. Interface dependency. The system becomes the lens. It's if it's not the surface, it doesn't exist. If it does not surface outside the system, so you got to know your business. It doesn't exist, really. If people don't know that it's it doesn't surface because the system hadn't brought it there. This is not the time to be lazy. It's never been a time to be lazy or lax of gazeable or whatever word you want to use for it. This is not the time, and it's never been the time. We have to inspect what we expect. And once that happens, authority is no longer where you think it is. Leadership decisions moments that we have to think about. I want to bring this into uh uh another real decision. I use my my my situation as one, I use the HR uh information as two. I started with the HR and then used me. So let's bring this into a real decision now. Uh, a leadership team reviews a strategic dashboard. The AI highlights a new market opportunity. The data looks strong, recommendation is clear, the team discusses it, debates it, approves it, right? But here's the question no one asked. Is this real? Did we discover this opportunity? What opportunity are you talking about? Or did the system define it for us? Because if the system framed the opportunity that you discovered, it shaped the decision, right? You decided based on what the system discovered, what the opportunity was. And if it shaped the decision, then therefore it holds the influence. Yeah. So what leaders must do with this information is embrace yourself. Because now here's the important part, guys. The goal is not to stop authority migration. That's not realistic. It's just not gonna happen, and it's not necessary. AI should influence decisions that that's where its value is. The goal is to govern it deliberately, okay? That means mapping where authority actually lives, defining decision rights clearly, designing governance structures that track influence, and reviewing it regularly. Because unmanaged authority migration is where risk concentrates. And this happens in real life with uh processes. This is real too, but this happens with processes where people, when things have to go through several different people, and they decide what they're gonna bump up to leadership, not including everything that was said, only what they want to bump up to the next level or tier of leadership. And a decision gets made, and you sit here thinking that's not what we said, that's not what we said back here when we were pulling that process together. Oh my gosh, this happens in situations in companies where we may not know what the real this future is and what they're uh moving towards. So in this day and time, you're and get this, as you're going through, people say, Well, you don't need a college degree because AI will have all the knowledge. AI don't know people, AI don't understand um things that are not uh written or embedded in the system. It can calculate things and give you a structure, give you an estimation on something, but to really understand something, it you have to know the process. And that's why when we do these training sessions, AI certifications, no matter what it's for, a manager, a process, uh CFO education, whatever the case may be, we ensure that that training has embedded in it the way a person, what that person would actually have to do. And they have to show us that they really know what they're talking about. And in their examples and fixing something that's in their organization right now that would need to be fixed, and building something for it, taking it back to their manager and telling us, is that does that work or does it not? What happened? Why, why are you using it? What are the risks you see involved in it? Yes, people must understand and know that. So when you have somebody on your team that's just flowing through, that's not good these days. You have to have people that are knowledgeable about what's happening in the organization, especially when you have AI at those beginning level runs, and that's moving things up the panel. And you needed that anyway, even if you didn't have AI in that process. So let's bring this back. Authority didn't disappear, guys. It migrated from leaders into systems that shape how leaders see the world. And the real question is not AI making decisions. The real question is, are you governing where authority lives? Okay, because if you're not, then the system in and of itself is yeah, the systems you build is making those decisions for you. So before the next episode, do this exercise. You're like, Dr. Carol, are you taking us through a class? I'm trying to build you into something. If you take my advice on these exercises I give you, you in and of themselves, you would say, Oh my gosh, I finished that episode, the podcast, all 12 series. And boy, did I learn a lot and take away a lot from it. Okay. You may call me back and say, Dr. Carol, uh, is there a certificate for this? Is there some kind of little exam or something like that or exercise I can do for you to show that I participated in your podcast? Because those exercises or the information you send back, I'm going to check it and I will know whether you understand AI or not. Not because you cheated andor may have gone to AI and said, How do you do this? I need you to know where your own data comes from. And the AI can't tell you that. And what options are being surfaced and which ones aren't in your organization? Uh who reviews the output in your organization, and who makes the final call in your organization? Sound sort of like a government. Uh the last one is who is accountable for the outcome? That's five questions. Where does the data come from? One, what options are being surfaced, and which aren't? That's two. Who reviews the output? Three. Four, who makes the final call? And five, who is accountable for the outcome? Then ask one more question. Where has authority already shifted? Because once you see it, you can't unsee it anymore. You know and you understand it. So as we bring this episode to a close, we take this one step further in our next episode. Because when decisions go wrong, authority migration creates something even more dangerous: an accountability vacuum. And most organizations are not prepared for it. No, they're not. So I'll see you in the next episode of Two Doctors and a Twist. Take care.