Divergent Compute.AI Economic Think Tank

The Index

The AI Fragility Index

One reading of how fragile the AI build-out is, on a 0–100 scale — the convergence of six filing-sourced indicators across the 43 companies that are the build-out. Higher = more fragile. It moves as the filings do.

/ 100
0 calm · 100 fragile
Open the screener → The full brief →

What's driving it

By layer of the build-out

Most fragile names right now

Methodology. Each company's composite is the weighted mean of its six indicator scores — depreciation integrity, capex-vs-demand, insider selling, circular financing, energy, and organic demand. The Index is the equal-weighted average of those composites across the build-out core (Layers 1–4: compute & infrastructure, hyperscalers & cloud, model labs, AI software). The broader-market comparators in Layer 5 are excluded — they're the control group, not the build-out. It's reproducible from the published company scores, and it sits alongside the divergence gauge, which reads the market-vs-fundamentals timing. Scores are internal judgment, read for convergence; figures sourced or labelled.