Skip to the album
Aminahstore

First day cover ·

AI Valuations Are Running on a Faster Clock Than Enterprise Buyers

Budgeting & Saving Posted by [email protected] (John O'Connell, MBA) 340 reads

AI models ship every few weeks, but regulated enterprises take months to buy them. That gap could matter more to AI valuations than the next model launch.

AI Valuations Are Running on a Faster Clock Than Enterprise Buyers

Model makers now ship new AI systems on a schedule measured in weeks. A bank, hospital, factory or retailer doesn't change its core technology on that schedule, and it probably never will. That difference looks like an operational footnote, but I think it's a structural risk, and it sits underneath the share prices of many companies in the AI supply chain.

Two clocks running at different speeds

Model providers are competing on speed. Release cadence compressed sharply through 2025 and into 2026. Labs that once shipped a major model about every six months were putting out multiple substantive releases a week by early 2026. One industry tracker counted more than a dozen meaningful frontier launches in the first quarter alone. A separate count logged five major launches or availability milestones in a single 16-day window in July 2026.

Buyers are running on a much slower clock. A 2026 analysis of AI vendor contracts found that legal review for AI-specific agreements now averages close to 15 weeks. That's more than three times the review time for a traditional Software-as-a-Service (SaaS) contract a few years ago. Add security review, compliance sign-off and a group of stakeholders who each need to approve, and a large enterprise deployment can take nine to 10 months from first evaluation to signed contract. That's true even when the vendor already holds standard security certifications.

Why enterprises move slowly, and why that's sensible

Slow adoption isn't a sign that big companies are stubborn. Every industry that handles sensitive data or safety-critical work has its own version of the same brake:

  • A bank or broker-dealer has to show examiners that its data governance and vendor oversight hold up.
  • A hospital system has to show privacy auditors how patient data moves through any new tool.
  • A drugmaker has to validate any system that touches a clinical trial or a manufacturing line.
  • A manufacturer has to clear safety certification before a new system goes near a production floor.

None of these checks runs on a six-week cycle, and none of them is optional for the companies running them. Consumer-facing businesses feel a lighter version of the same friction. A retailer adding an AI tool to customer service or logistics still has to test it against real order volume, train staff and confirm it won't break during a peak season. The specific obstacle changes by industry, but the pace stays roughly the same.

The gap shows up in the numbers. Deloitte's 2026 State of AI in the Enterprise report found that worker access to AI tools rose 50% in 2025. Yet the share of companies with at least 40% of their AI projects in production is only on track to double within the next six months, which leaves most organizations well short of that mark. My read is that most of this gap is a governance and procurement problem rather than a technology one, and it won't close on the timeline model providers are shipping against.

The assumption this breaks in the supply chain

Every layer of the AI supply chain, from chip design through data center construction to hyperscaler capital spending, rests on a bet about how quickly enterprise customers will adopt and pay for AI. Combined 2026 capital expenditure from Microsoft, Amazon, Alphabet and Meta is projected at $700 billion to $725 billion, according to a Yahoo Finance report. That spending only pays off if enterprise revenue arrives at a pace that matches how fast the infrastructure is going up.

Here's the thing: if enterprises need 12 to 24 months to take one AI deployment from evaluation to production, and model providers keep shipping new frontier models every few weeks in the meantime, the two clocks aren't just out of sync. They're structurally incompatible. Infrastructure gets built for a demand curve that assumes fast adoption, while actual revenue follows a curve shaped by compliance review, legal sign-off and board-level risk tolerance.

Where the exposure lands first

The companies most exposed here probably aren't the chip designers or the data center builders. Those firms get paid when hyperscalers buy equipment, whether or not that capacity gets resold quickly to enterprise customers. The pressure is heaviest on enterprise software companies that sell AI-enabled products into regulated industries. They depend on enterprise contracts to turn infrastructure spending into revenue, and those contracts are the slowest part of the chain.

There's a tension built into the software pitch, too. A vendor whose selling point is rapid iteration is marketing against the clock its buyers are forced to run on. Fast model refresh is a real advantage for a consumer app competing on user experience. For a vendor selling into a compliance team that has to document, test and defend every material change before it goes live, the same speed can become a liability.

What to watch before the next earnings cycle

This kind of risk rarely arrives as one dramatic event. It tends to appear as a slow gap between guided revenue and delivered revenue at the enterprise software layer, quarter after quarter. Meanwhile, infrastructure companies keep reporting strong numbers because hyperscalers keep buying equipment. A rising capex forecast tells you demand for infrastructure is still strong. It doesn't tell you whether the enterprises that will eventually pay for that infrastructure can move fast enough.

If you follow AI-exposed companies, these are three general places to look:

1. Read the backlog language, not just the revenue line, for enterprise software companies selling AI features into banking, healthcare or manufacturing. A growing backlog of signed but unrecognized contracts shows that demand exists but hasn't cleared procurement yet. A shrinking backlog next to flat AI revenue would be an early sign that the sales cycle is running longer than guided.

2. Compare guidance language from one quarter to the next. If a company has described AI revenue as "accelerating in the back half" two quarters in a row, it's fair to ask why that acceleration keeps getting pushed out. Procurement timelines don't compress on command, so guidance that assumes they will deserves a closer look.

3. Think about infrastructure and enterprise software as separate exposures. Chip and data center companies are tied to the hyperscaler capex cycle, while enterprise software companies only get paid when adoption actually happens. If the two sit side by side at the same weight, you're effectively pricing them on the same clock, and they aren't on the same clock.

The market is pricing AI companies on an innovation cycle measured in weeks. Most of their eventual customers move on a cycle measured in years, and a new model release doesn't close that gap. If a market correction does come, I doubt it will be caused by AI failing to work. It's more likely to be triggered by investors finally pricing in how long enterprises actually take to pay for it.

Source: [email protected] (John O'Connell, MBA) · www.kiplinger.com

From the same series