The AI Decision-Rights Map Every VP Needs Before the Next Steering Meeting

The five questions every RevOps VP should be able to answer about each AI system before the next steering meeting.

Before your next steering meeting, you should be able to answer five questions about every AI system you run. Most revenue operations VPs can’t answer them for even one. Not because they don’t know their systems — they know them cold. Because the questions aren’t about the system. They’re about authority. And the authority was never handed out.

The five questions

Here they are. For each AI system you’re accountable for: Who decides when it’s wrong? Who funds it? Who can say no? Who owns the number it produces? Who carries the risk when it fails?

If you can’t put a name to each one — not “a committee,” not “we,” a person — that’s not a gap in what you know. It’s a hole in the structure. You’re accountable for a system nobody governs. No one can stop the AI, because no one was given the right to. And the blame for that hole sits on you, because you’re the one who understands the machine well enough to take it.

The real question under the map: are you authorized to govern the revenue stack you're accountable for? Or are you absorbing the risk of systems you can't actually govern — with the decision rights of someone who just runs a function?

Walk one system through the five

Take the AI model that scores and routes inbound leads. Walk the five, honestly.

Who decides when it’s wrong? The model misroutes, and a key account sits untouched for two weeks. Who can say “the model is wrong, override it”? Usually no one. You notice first. Noticing isn’t authority. So the call waits, or it becomes a workaround.

Who funds it? Finance paid for it. But renewals run on autopilot. The tool is “essential” because it’s wired in, and no one with budget power is actually checking whether it should stay. That’s funding without governance.

Who can say no? Sales wants to add another AI layer to the routing. Who can refuse on how it’s built? The person who understands how it’s built — you — usually can’t say no. You can only warn. A warning isn’t authority either.

Who owns the number? The model produces a score that drives pipeline. Someone questions that number in a QBR. Whose number is it? Yours, because you’re closest to it. You didn’t design the model’s logic. You own its output in the room.

Who carries the risk? The model produces a biased or non-compliant result. Whose name is on it? Yours. Always the builder’s.

Five questions, one system. The honest answers: no one, no one, you-but-with-no-power, you-because-you’re-closest, you-by-default. That’s the hole. You’re standing in the middle of it.

Why the map keeps coming up empty

This is the flaw under most “AI governance” efforts — and most revenue operations consulting. They map who’s involved. They never assign who decides. Involvement gets written down. Authority doesn’t. A chart with four names under “consulted” and an empty box under “accountable” isn’t governance. It’s the paperwork of governance wrapped around the same hole. And because the rights were never assigned, the decisions keep flowing to whoever’s closest — by accident, not design.

The American Management Association found in 2026 that 69% of executives spend at least half their time influencing people they have no authority over. That number is this exact hole, measured across the economy: accountability sliding sideways and down, onto people with responsibility and no power.

Accountability by closeness isn’t authority

It’s worth being exact about how this works, because it’s the part the company keeps misnaming. When a decision right isn’t assigned, the decision doesn’t stop. Systems decide all day. It just routes to whoever’s closest and capable enough to be blamed. In a revenue org, that’s almost always the person who built the system.

That’s accountability by closeness. From the inside, it feels like ownership. That’s the trap. You’re in every meeting about the system. You’re the one who gets paged. Your name is on the invite. None of that is authority. Authority is the right to say no and have it stick. The right to stop the thing and not get overruled by someone who doesn’t understand it. Closeness gives you all the exposure and none of the standing. And because it looks like ownership, you can carry it for years before you notice it was never yours.

What the empty boxes are telling you

When you build the map and the boxes come back empty, the urge is to fill them with your own name. You’re already doing the work, so you write yourself in five times and call it ownership. Don’t. An empty box isn’t a job opening for you. It’s proof the company put a system in place without deciding who governs it. Writing your name in finishes the defect instead of fixing it.

The boxes that should hold someone else’s name — the CRO who owns funding, the risk or legal owner who carries compliance, the executive who can authorize a stop — are the company’s to fill. The map’s real value is that it makes those empty boxes impossible to ignore, in a room with the right people in it. You’re not building yourself a to-do list. You’re building the document that ends the silence.

What this costs

Grant Thornton's 2026 AI Impact Survey put a number on it. Three in four boards have approved major AI investments, but only 52% have set clear AI governance expectations. The map exists to make you one of the few who can name an owner — at least for the systems with your name on them. Because when governance is unset and the system fails, "unassigned" turns into "the builder." The map changes the answer before the failure forces it.

What changes when the rights land

It’s worth being concrete about what shifts when the five rights get assigned. The change is bigger than it looks on paper.

The day a named executive owns the stop decision, the model stops being your risk. It becomes the company’s system, run by someone with the standing to run it. The day funding sits with a person who’s actually checking the tool, renewals stop running on autopilot, and the stack stops piling up cost no one chose. The day someone else can say no on how it’s built, you stop being the only voice warning about the next layer. The “no” has an owner now, and the owner has authority.

None of this asks you to be more persuasive. It asks for the lines to be drawn and signed. That’s the whole difference between carrying a system and governing one. It’s a difference in structure, not effort.

What fixed looks like

The map isn’t the fix. Filling it in with real names is — and changing the boxes where your name sits in a spot that should hold someone else’s.

I’ll give you the map, because the five questions are worth more in your hands than mine. What I won’t hand over in a blog post is how to score the gap, how to decide which holes to close first, or how to order the work for your company. That’s the actual work, and it’s specific to you. A rev ops advisor takes this map — the Executive Authority Method — finds every place your accountability runs past your authority across the AI stack, and puts the fixes in order.

The move

Build the five-box map for your single most important AI system. One row. Five owners. Bring it to the steering meeting and ask the group to fill the empty boxes. The blanks aren’t yours to absorb. They’re the company’s to assign — and asking the room to assign them is the most structural thing you can do in sixty seconds.

Run the Checklist.  The Executive Authority Gap Checklist to locate your authority gap across the whole stack — every place you’re accountable for outcomes you can’t govern →

Sources:

Grant Thornton 2026 (52% boards w/ clear AI governance), AMA 2026 (69%)