For years, you didn’t really manage your cloud commitments. You outsourced the risk — and crossed your fingers. 

Let’s be honest: as IT leaders, we’ve trusted vendors to smooth usage patterns across customers. We’ve believed in marketplaces to create exit options for excess capacity. We’ve nodded along when engineering assured us they’d “optimize away” any overages before they hit the books. We’ve operated in a world where massive financial commitments somehow didn’t feel like actual commitments. 

That era is over. AWS now says: “If you sign a commitment, you own it.” Full stop. 

The June 2025 policy change ending commitment pooling and discount sharing across customers isn’t just an adjustment to partner terms. It’s the abrupt end of a collective fantasy that cloud spend was manageable by someone else. AWS has called in our tab, and now we have to pay the bill. 

The Fantasy of Flexibility 

The illusion was elegantly constructed. For years, managed service providers and resellers performed financial sleight-of-hand that made our cloud commitments feel like discounts rather than liabilities. They pooled Reserved Instances (RI) and Savings Plans (SP) across hundreds of customer accounts. If your usage dropped, no problem — someone else in the pool would take up the slack. Sub-accounts were reshuffled like deck chairs to match shifting usage patterns. RI marketplaces made commitments feel tradeable and liquid. 

Companies like DoiT International, Zesty and Spot.io built business models on commitment arbitrage—aggregating buying power and spreading the risk. Startups like Pump openly promoted “group buying for AWS commitments,” effectively giving small companies the leverage of an enterprise. We grew comfortable in this cushioned reality. 

But these weren’t strategies. They were financial conveniences that worked exactly until the moment AWS decided it was done subsidizing the flexibility. 

AWS’s policy shift draws a line in the sand: Commitments are now non-transferable, non-shared and strictly enforced. The memorandum is clear: Don’t commit to what you can’t use yourself. 

Why AWS Pulled the Plug 

This wasn’t a capricious move by AWS. It was an inevitable correction driven by the cloud provider’s own economic imperatives. 

AWS is not just a service provider — it’s a capital-intensive infrastructure business with a staggering $100 billion annual capex footprint. They build data centers years before you rent space in them. They provision hardware months before your workloads run on it. And they need reliable demand signals to justify these investments. 

As Andy Jassy put it, AWS must “procure data center capacity ahead of when we monetize it.” When vendors abstract away commitment risk, they undermine AWS’s planning model. If a cloud reseller has aggregated $100 million in customer commitments but those customers aren’t actually individually committed, what happens when usage patterns shift dramatically? The commitment becomes a fiction—useful for discounts but useless for capacity planning. 

This isn’t about punishing vendors or squeezing more margin (though AWS won’t complain about the potential windfall). It’s about protecting cloud economics at a global scale. When thousands of companies can reduce their commitment risk through third-party arbitrage, AWS loses its ability to predict demand accurately. And when the system stops working for the platform itself, change becomes inevitable. 

What Companies Got Wrong 

Many IT and finance teams treated cloud commitments as pure savings vehicles, not the financial liabilities they actually are. We believed vendors had contingency plans for underutilization. We assumed cloud was inherently elastic—until it wasn’t. 

Most critically, we conflated the flexibility offered by vendors with the flexibility inherent to AWS itself. But that flexibility was never an AWS feature—it was a vendor artifact, a layer of risk intermediation that AWS has now eliminated. 

As a result, our forecasts were built on risk that wasn’t modeled. Usage volatility was swept under the rug. We calculated ROI on cloud migration without factoring in the true cost of commitment management. Cloud finance looked better on PowerPoint than it performed in practice. 

Even sophisticated CFOs would often delegate cloud financial management to IT or procurement, assuming that a PO and a discount tier were sufficient governance. We let cloud spending bypass the rigorous financial controls we apply to other major expenditures. We convinced ourselves that the cloud’s consumption model somehow exempted it from standard financial discipline. 

“We’ll grow into it,” became the blanket assumption covering shaky commitment forecasts. “The vendor will sort it out,” became the fallback when usage didn’t materialize as planned. These were never sustainable approaches to financial management—they were postponements of financial reality. 

What Happens Now 

The risk hasn’t disappeared. It has simply reverted to where it always belonged: Your balance sheet. 

If your company has made commitments based on pooled models, you may be severely overexposed. If your usage drops, you own the delta—and there’s no one to bail you out. Some vendors are literally holding nine-figure obligations they can’t fully utilize after June 1. That’s not a rounding error; it’s an existential threat. 

The potential fallout is sobering: Stranded commitments that still must be paid, vendor insolvency as resellers struggle with unfulfilled obligations, and unexpected margin hits for companies that assumed their cloud costs were optimized. 

The concerning reality is that many CFOs don’t have clear visibility into which of their discounts are direct AWS relationships versus those that depend on now-banned mechanisms. Do you know what happens when your vendor’s model breaks? If a FinOps vendor promised you “automated commitment management” or “dynamic optimization,” how exactly did they deliver that? If the answer involves cross-customer resource sharing, that capability vanishes on June 1st. 

Even more troubling is the timing. Many cloud commitments are three-year terms, meaning commitments made in 2022-2023’s cloud exuberance are still in effect, while actual usage may have rationalized. Some resellers report utilization rates below 60%—with no relief in sight. 

The Maturity Moment 

AWS’s move is painful—but clarifying. This is the moment when cloud spend finally graduates to real finance. Not a variable utility. Not a “discount optimization” line item. Not an engineering concern with a financial footnote. Cloud is now a structured financial asset that requires real governance and oversight. 

Smart finance and IT leaders are re-auditing their commitment exposure—both direct and through partners. They’re building worst-case scenarios where discounts disappear overnight. They’re demanding transparency from their vendors about post-June viability. And they’re rethinking how cloud fits into their cost model when the risk can’t be offloaded. 

This moment echoes other financial coming-of-age stories in tech: The shift from perpetual licenses to SaaS revenue recognition, the evolution of data center infrastructure from capital to operational expenditure, and the maturation of cybersecurity from an IT concern to a balance sheet risk. 

Companies that embrace this transition will gain what they’ve always wanted: Actual control over one of their largest cost centers. The tradeoff is that control comes with responsibility. If you overprovision, you pay the price. If you underestimate growth, you pay on-demand premiums. There is no middleman to absorb the variance anymore. 

Cloud Risk Isn’t Theirs Anymore, It’s Yours 

You can’t outsource this anymore. This is your infrastructure. Your forecast. Your risk. 

Cloud has become a fixed cost with volatile utility—which means it must be modeled, governed and owned at the finance level just like debt, leases, or any other financial obligation. The shell game is over, but now you can lead with discipline. 

AWS didn’t pull the rug. They just turned on the lights. And what’s revealed isn’t pretty: We’ve been treating multi-million dollar financial commitments with the same rigor we apply to office supplies. We’ve delegated core financial governance to vendors with misaligned incentives. We’ve built forecasts on assumptions of flexibility that weren’t contractually guaranteed. 

The good news? IT and finance leaders now have both the imperative and the opportunity to bring cloud spending under proper financial stewardship. The clarity is worth the pain. 

The clouds haven’t changed. Only the illusion of who bears the risk has been dispelled. It was always yours. Now you know it.