Your AI Agent Bill Is Unpredictable Because Your Agent Is

Last week I caught up with a friend. His company just green-lit a rollout of both Microsoft Copilot and Claude across the whole org. Big win on paper.

He wasn't celebrating. The cost was part of it, sure. But the real weight was the governance and compliance overhead the new billing model dragged in with it. Who can use what. Which users get capped at which tier. Which department cost center eats the consumption. How you even audit a bill that moves every month.

Microsoft didn't create the problem. Consumption pricing simply put a meter on it. The bill isn't telling you your AI is expensive. It's telling you which workflows were expensive all along.

What Microsoft actually changed

On June 16, Microsoft launched Copilot Cowork, an agent that runs office tasks on its own. Drafting documents, building spreadsheets, comparing files, prepping meetings. The tool still needs a paid Microsoft 365 Copilot subscription. But now every task it runs is billed separately, based on how much compute it burns.

The unit is a Copilot Credit. Pay-as-you-go runs $0.01 per credit, or you prepay packs of 25,000 credits for $200 a month. A single agent response can eat several credits depending on how much it retrieves and does. Cowork itself runs on frontier models: Opus 4.8, Sonnet 4.6, GPT 5.5 on the top tier.

To keep the bills from running away, the service ships disabled by default. Admins cap spending per employee, per team, or per department.

Charles Lamanna, Microsoft's executive vice president for Copilot and agents, described the shift plainly to AFP. It's “like you're filling up your gas tank at the pump.” His reasoning: there's no single user license that makes sense anymore, because different users consume wildly different amounts of compute.

Microsoft moved off pure seat licensing because AI workloads vary dramatically by user and by task. That's a fair reason. Consumption pricing aligns price with compute, and when compute genuinely varies, that's the honest way to charge for it.

What consumption pricing actually exposes

Here's the part worth sitting with. Consumption pricing makes sense when compute usage genuinely varies. The challenge is that many enterprise workflows don't vary nearly as much as the pricing model assumes.

A Microsoft rep would tell you, correctly, that Copilot was never meant to replace deterministic workflow automation. Agree with that. The problem was never Microsoft's pricing. Consumption pricing is just a scale that reads out a number you were already generating. What it exposes is which workflows should never have been implemented as generic agents in the first place.

A general-purpose agent can't assume anything about your workflow. Every run starts over. It retrieves context again, reasons again, and spends compute rediscovering the same process. That's valuable for open-ended work. It's unnecessary for deterministic work. You were paying for the inefficiency all along. Consumption pricing simply made it visible, one department and one cost center at a time.

That's what my friend is fighting. Two teams with the same headcount produce very different bills because one runs retrieval-heavy workflows while the other mostly asks simple questions. Finance can't reliably forecast a number that changes every month for reasons nobody fully controls.

Then comes the governance overhead: spending caps, tier assignments, cost-center routing, documentation, audit trails. The tool is metered. The overhead of managing the meter isn't.

Caps don't control cost. They ration value.

Here's the trap in the cap mechanism. Finance sets a per-user or per-department ceiling to make the bill safe. Fine. But that ceiling is measured in compute, not outcomes. When a user hits the cap, the agent doesn't finish cheaper. It stops. Mid-task.

So the cap doesn't control cost. It rations value. The finance-approved ceiling becomes the exact point where the tool quits being useful. You've made the bill predictable by making the usefulness unpredictable. That's a bad trade, and it's the only trade a metered generic agent offers.

The better approach: build the system, don't rent the agent

The fix isn't a cheaper agent. It's a different architecture. You stop paying to re-buy generality on every run and pay once to engineer the workflow.

• Deterministic code for the boring 80%. Most of what an agent does isn't fuzzy: parsing, formatting, moving data, conditional branching. That's code, not AI. Write it once and it runs for effectively nothing, the same way every time.
• Right-size the model for the 20% that's actually hard. The expensive anti-pattern is a frontier model doing trivial work. Route classification and extraction to a small, cheap model and reserve the frontier call for genuine judgment.
• Engineer the context. Agent loops re-send a growing transcript on every turn, so a ten-step task costs far more than ten single calls. Compact state between steps and retrieve only what the next step needs.
• Batch what isn't realtime. Most “cowork” work isn't synchronous. Overnight reports, bulk processing, and scheduled summaries run async at roughly half the price.
• Narrow pipelines over one do-everything agent. A general agent burns tokens exploring because you told it nothing. Decompose the job into narrow flows with model calls at specific decision points, and save the agent for genuinely open-ended work.

The payoff is the thing the metered agent structurally cannot give you: a predictable cost per run. When the workflow is fixed, the system does the same work every time, so it costs the same every time. You can put a real number in a cost-center forecast and defend it.

The pattern under all of this

We've been saying AI sprawl is content sprawl with a new logo. This is the same disease in a new organ. Call it orchestration sprawl. Companies are bolting general-purpose agents onto every workflow, paying a metered frontier model to rediscover the same process thousands of times, and then capping the bill until the tool breaks.

Generic metered agents have a place. For exploratory, one-off, unpredictable work, gas-pump pricing is honest and fine. Pay for what you use when you don't know what you'll use.

For a repeated production workflow, it's the wrong tool. You know exactly what that workflow does. So you engineer the cost out of it once, and you own a system with a bill you can predict instead of renting an agent with a bill you can only cap.

This is why we increasingly design AI as one component of a larger system rather than the system itself. The model should provide judgment where judgment is valuable. Software should handle everything else. That split is what makes a workflow predictable, auditable, and cheap to run, and it's the opposite of handing the whole job to a generic agent and metering the result.

The caps, the tiers, the cost-center routing, the audit trail my friend now has to maintain: all of it is governance overhead he inherited because the workflow lives on a meter. Move the workflow into a purpose-built system and most of that overhead evaporates. Fixed behavior, fixed cost, one thing to document instead of a consumption model to police.

The rollout will work. But the companies that win with enterprise AI won't be the ones with the biggest agent budget. They'll be the ones that know where not to use one. Because what looks like a licensing problem is usually an architecture problem wearing a licensing costume.

Generic agents belong where the work is unpredictable. Your business processes shouldn't be.

Sources

Pricing and product details in this post are drawn from the following, current as of publication. Microsoft's consumption pricing moves, so confirm live figures before acting on them.

• Charles Lamanna quote, June 16 GA date, pay-as-you-go mechanics, spending caps: AFP, via Dawn
• Copilot Cowork models (Opus 4.8, Sonnet 4.6, GPT-5.5), launch date, Microsoft 365 Copilot requirement: Let's Data Science; Project Flux
• Copilot Credit pricing ($0.01/credit) and 25,000-credit packs at $200/month: Microsoft Copilot Studio pricing page

Matt Fishman is Founder and CEO at Privion. We design AI systems where the model provides judgment and software handles everything else: fixed behavior, predictable cost, and a bill you can put in a forecast. More at priviontech.com.