Something changed in the market this month, and it will reach your Azure estate before the quarter is out.
Alphabet reported. Google Cloud revenue was up 82% year on year to $24.8 billion. The stock got sold anyway, because the same release guided 2026 capital expenditure to roughly $205 billion and showed free cash flow going negative for the first time in the company's life as a public business. A beat on revenue was not enough to offset the spending line. For about two years, "we are building more AI infrastructure" was a reason for a stock to go up. As of this week it is a question the company has to answer. Microsoft and Meta report tomorrow after the close, Apple and Amazon on Thursday, and all of them have the same question waiting.
The repricing is narrow
Broad equity markets are close to flat this week; the selling is in AI infrastructure suppliers. Nvidia is down 5%, AMD 5.2%, Micron 2.3% and Sandisk 11% in a single session. That follows a worse stretch in early July, when Intel dropped 21% across the selloff. Reports said SK Hynix was slowing its high-bandwidth memory expansion, and Meta said it had surplus AI infrastructure capacity worth monetising, after two years of telling everyone it could not build fast enough.
The selloff suggests the market is questioning whether demand for AI compute will keep outrunning supply for long enough to justify the build cost. Interest rates are the other half. The Fed has held rates at 3.50-3.75%, and its own projections shifted this quarter from expecting a cut by year end to implying a rise. Expensive money makes long-payback investment programmes get scrutinised.
How it reaches you
Public markets stop rewarding infrastructure spend on faith. Boards read that, CFOs read the boards, and within a quarter or two someone senior asks a version of this:
What are we actually running, why, and what does it return?
A cloud bill tells you what you were charged. It does not tell you what is running, who asked for it, whether it is compliant, whether anyone uses it, or whether it can be turned off without breaking something. The only function in the business that can answer all that is the one that defines the estate in code, which in most organisations is the platform team.
For a platform team, a capex reckoning arrives as a governance and visibility problem long before it arrives as a budget cut. You get asked to justify 100%, not to trim 20%, and the honest answer for most estates is "I would need a few weeks to find out."
What shipped in July
HashiCorp's release slate reads like a direct answer to that question. I do not think that is an accident; vendors build toward where budget pressure is heading.
tfpolicy (announced 16 July, public beta on HCP Terraform) is policy as code written in HCL, the language your infrastructure is already defined in, rather than Sentinel or Rego. It evaluates resource relationships, external data, provider and module usage, and live infrastructure after deployment. The last one matters most. Evaluating real state, not only the plan, is the difference between "we block bad Terraform" and "we can tell you what is non-compliant in production right now".
Sentinel and OPA are still supported, so nothing needs ripping out. It is still a public beta, so pilot it on a non-production project and write three or four policies that matter before you plan a migration.
Terraform Stacks became adoptable on estates that already exist. The batch that landed on 23 July includes:
- Monorepo support (GA), so several Stacks can share one repository without every commit firing every run. If you rejected Stacks on repo-layout grounds, that objection is gone.
- A guided CLI migration from workspaces to Stacks (public beta), the actual blocker for most people.
- Workspace and Stack restore (GA on Standard and Premium), with deleted items held for 30 days.
Infragraph (HCP Terraform, limited availability for Standard and Premium on US instances) now has connectors for Azure, GitHub and Terraform Enterprise. A graph across the estate is what you need when someone asks what depends on what, and before you can safely turn anything off.
The Azure homework
azurerm 4.81.0 shipped on 14 July. It fixes a CDN Front Door wildcard custom domain validation regression that 4.80.0 introduced two weeks earlier, which is your periodic reminder to pin the provider and read the changelog before you bump.
Microsoft's retirement wave is a cost conversation with a deadline. NVv3 and NVv4 support ends 2026-09-30. Several legacy Reserved VM SKUs stopped selling on 2026-07-01. CuRP retires 2026-10-31, and HBv2, HC and NP series go on 2027-05-31. If your estate has NVv3 or NVv4 in it, that is a migration for the next two months, and far better done as a planned modernisation than as an emergency in September.
Patch OpenTofu
OpenTofu 1.12.5 and 1.11.13 landed on 21 July, both with a security fix. The Encrypted Client Hello implementation was leaking pre-shared key identities during the handshake, which let a passive network observer de-anonymise the hostname the client was contacting, which is what ECH is meant to prevent. Both maintained lines carry the fix, so there is no reason to sit on it.
What Terraform core is cooking
Stable is 1.15.8. The 1.16 alphas (most recently 1.16.0-alpha20260715) include import blocks inside modules, which has been awkward for brownfield landing zone work because imports could not live where the module lived. Don't pin production to the alphas, but do read the changelog. The rest is in this month's round-up.
What to do this week
- Patch OpenTofu to 1.12.5 or 1.11.13 if you run it.
- Find your NVv3 and NVv4 instances. Knowing the number now is worth more than a plan you write in three weeks.
- Write down what you would say if a CFO asked what your estate costs and what it returns. Write down the process you would use to answer, not the answer itself. If the honest version is "I would need two weeks," that gap is the finding.
- Pilot tfpolicy on one non-production project. Three policies, one of them evaluating live state, because that is the capability your current gate lacks.
- Re-examine Stacks if you dismissed it. The monorepo and migration objections were real, and this batch answered them.
The point
Whatever the market does next, scrutiny of infrastructure spend travels downward until it lands on whoever can produce evidence about the estate, which in most organisations is a platform team with a Terraform repository and no easy way to answer questions about live state. The tooling to answer those questions shipped this month, and the question arrives next quarter.
Sources: HashiCorp Terraform blog · tfpolicy announcement · Stacks and workspace restore · azurerm releases · OpenTofu releases · Terraform releases · Trading Economics, US markets · Forbes, the July semiconductor selloff · Kiplinger, July Fed meeting · Azure updates
Share price moves quoted as at 28 July 2026. Nothing here is investment advice.