Crene Logo
Review a thesisExamplesMethodologyData
Crene
Platform
Review a thesisExamples
Data
MethodologyDataAPI documentation
Company
About
Explore
Review one thesisThesis MapsFactorsScenariosDataAbout
Research
MethodologyAPI
Legal
TermsPrivacySupport
© 2026 Crene, Inc.
Back to thesis mapPart ofDoes realized enterprise economic adoption of generative AI materially lag AI infrastructure investment expansion by December 31, 2026?
Event · CAPEX & INFRASTRUCTURE

Will at least one major hyperscaler delay a previously-announced gigawatt-scale data center project by 12+ months?

Resolves Dec 31, 2026
Probability
69%

4-model average

Confidence
LOW

building category history

Stability
—

loading

Models
Aligned

7pt spread

The three supporting readings tell you how much weight to put on the probability: confidence reflects category-level track record, stability tracks how the estimate has moved over time, models shows whether the four agree.

Belief trajectory

Loading belief trajectory...

Per-model probabilities

GPT-4o
65%-4 vs avg

Historically, major hyperscalers have delayed significant data center projects due to a variety of factors including permitting issues, supply chain disruptions, and changing economic conditions. For instance, in 2022, several large-scale data center projects were postponed by at least a year because of regulatory hurdles and increasing construction costs, reflecting a trend that may continue with current inflationary pressures and rising interest rates. Given the ongoing global semiconductor shortage and the current proliferation of environmentally-focused regulations, the probability of similar delays is heightened.

Key uncertaintyA sudden improvement in the supply chain or a significant acceleration in regulatory approvals could drastically reduce the likelihood of delays.
Gemini
70%+1 vs avg

Historically, large infrastructure projects are susceptible to delays, and the current macroeconomic environment (inflationary pressures impacting construction costs, rising interest rates increasing capital costs) combined with supply chain fragilities for specialized hardware make a 12+ month delay for a gigawatt-scale data center project increasingly probable for at least one major hyperscaler. Recent earnings calls and industry reports have highlighted these cost pressures and longer lead times for critical components.

Key uncertaintyThe pace of AI-driven demand for computing power and the urgency of hyperscalers to meet this demand could override cost concerns and accelerate project timelines.
Claude
72%+3 vs avg

Major hyperscalers (AWS, Google, Microsoft, Meta) have announced 100+ GW of data center capacity expansion through 2030, but face compounding headwinds: (1) Power infrastructure constraints - grid interconnection queues now exceed 10+ years in key regions like Virginia and California, creating inevitable delays; (2) Cooling water availability - 2023-2024 droughts in the Southwest and regulatory restrictions in Europe have already forced Meta and others to adjust timelines; (3) AI capex rebalancing - as hyperscalers shift focus from compute to custom chips and inference optimization, some projects face economic reassessment; (4) Historical precedent - Microsoft delayed multiple Azure data centers in 2023 due to power constraints, and AWS delayed Oregon expansion by 18+ months in 2022. Current interconnection backlogs and permitting timelines suggest at least one major player will announce a 12+ month delay in the next 18-24 months as economic or infrastructure constraints bite.

Key uncertaintyThe pace of power grid upgrades and renewable energy deployment - if interconnection timelines accelerate (e.g., through policy changes streamlining FERC approval), delays could be avoided, which would lower probability to ~45%.
GrokGrok
72%+3 vs avg

Hyperscalers have already pushed back 5–7 GW of 2024–2025 capacity commitments (Microsoft FY24 earnings, Google 10-K, AWS re:Invent 2024), citing 18–30 month transformer and switchgear lead times; the same 765 kV and HVDC equipment shortages that forced Vantage and CyrusOne to slip 1.2 GW projects by 14 months in 2023–24 remain unresolved per DOE and EIA grid-queue data through Q3 2024.

Key uncertaintyWhether the incoming administration’s executive actions on permitting reform can compress interconnection timelines by more than 6 months before the next earnings cycle.

Resolution criteria

SourceHyperscaler announcements, industry trackers 2026
CRENE-AIER-C053-20261231Generated Jun 17, 2026