AI Is Not Eliminating Engineering Managers. It Is Separating Them.

Google cut 35 percent of its managers last year, transitioning those who managed three or fewer direct reports to individual contributor roles. Meta followed. The announcement language at every company was similar: AI now handles what managers were doing, so the layers are coming out.

The conclusion most people drew: managers are becoming obsolete.

The conclusion is wrong. The eliminations are real. Both are true simultaneously, and the gap between those two statements is where career durability lives or doesn’t.

What AI is actually doing is separating two roles that were always distinct but were often performed by the same person carrying the same title. The coordination manager, whose function was information relay, task tracking, status aggregation, and progress synthesis, is being replaced, because that specific work was always a symptom of a poorly designed system. AI executes it better, faster, and without the organizational drag that comes from routing everything through a single person.

The systems-architect manager is experiencing something different. Their function is to decide how human judgment, AI output, onshore capacity, and offshore execution combine into something that produces measurable results. They design the decision rights architecture that determines what requires a person and what can be delegated to a process. They build the feedback loops that improve team performance without their constant presence. When they are out, the team does not collapse. When AI becomes available, they extend it into the architecture rather than being replaced by it.

AI surfaced a distinction that was always there.

The gap between these two roles traces to a foundational problem in how management has always been taught. Management content is almost entirely person-centered: what the manager does, how the manager develops, which behaviors to practice, which internal work to do. Almost none of it asks whether the system the manager operates within is deliberately designed at all. Teams develop operating systems whether their leaders design them or not. Decision routing, escalation patterns, information flow, and trust accumulation form through habit and repetition. The manager who did not design those patterns is operating an accidental system. Accidental systems produce accidental results, and they center on the person holding them together by presence rather than architecture.

The question the person-centered framework never asked: could the team operate at a high level without the manager in the room? A team that cannot has a design problem beneath it.

There is a second pressure that is subtler and more immediate. The metrics that evaluate managers have not caught up with the work that actually creates value. The manager who redesigned how a team integrates AI output, who held fixed costs while reducing variable costs, who built reporting that ties team performance directly to revenue: that manager may be doing the highest-value work in the organization. The evaluation system often cannot see it, because it was designed for an earlier version of the role.

AI changes this equation. The reporting that makes systematic management work visible is now fast enough to build consistently: performance data, cost analysis, revenue attribution, and team output summaries that connect what a team produces to what the business needs. The manager who uses this data to control the narrative of their own contribution is operating in a fundamentally different competitive position than the manager who leaves their evaluation to systems that were not designed to see what they built. If you are not telling that story, someone else is telling it for you, using metrics that were not designed to capture the work that actually matters.

Here is the thing most managers will not say. I have watched technically excellent managers make this trade without knowing they were making it: being the person the team routes everything through feels like control. The Escalation Default, where decisions flow up because the system was never designed to hold them at lower levels, creates a constant sense that your presence is essential. It is. The system has a design gap at its center, one that compounds quietly while it still feels like security.

The manager who stays at the center because it is comfortable is building a cage, not a career. An organization that cannot function without you has no incentive to place you somewhere better. Your value does not transfer to the next employer, because it is not housed in architecture; it is housed in a specific relationship with a system you never designed to run without you. Undeniable business and revenue value transfers. Personal indispensability does not.

The diagnostic question: if you were absent for thirty days, would your team’s operating quality hold, degrade, or improve? The answer is not about your people. It is about the design of the system beneath them.

The managers willing to look at that question honestly, to see the system underneath their team, perhaps clearly for the first time, are the ones this moment will favor.

I write about structural leadership for technical leaders in high-stakes operating environments. The full operating model is in LeadershipOS™: http://TheLeadershipOSBook.com

---

I write about structural leadership for technical leaders in high-stakes operating environments. The full operating model is in LeadershipOS: http://TheLeadershipOSBook.com