OpenAI says 97.9% of its employees are now using Codex, up from ~40% in August 2025; non-developer usage of Codex has risen 137x for individual users
THE SO WHAT
Near-universal internal Codex adoption and a 137x jump in non-developer usage is a live case study in what happens when an org actually saturates on AI assistance. If your AI usage is still concentrated in engineering, assume your real upside—and process risk—is in operations, finance, and support adopting similar tools next.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIThe AI Trade Is No Longer About Owning One Thing: Taking Stock
The AI trade is fragmenting—theme intact, but picking the wrong node in the stack is now costly as dispersion between winners and laggards widens. Operators should stop treating “AI exposure” as a single bet and instead map where value accrues in their own value chain before aligning vendor and equity exposure.
Applied AISemiAnalysis Xie on Asia AI Supply Chain
SemiAnalysis highlighting specific Asian hardware suppliers as AI demand winners underscores how much leverage has shifted to a few upstream nodes. If you’re building on-prem or custom AI hardware, vendor concentration in Asia is now a strategic risk variable, not just a procurement detail.
Applied AIState Street: Asia May Win in AI Boom
State Street’s view that Asia may be the biggest AI beneficiary reflects where fabs, memory, and key components actually sit—not just where models are trained. For operators, that means AI hardware resilience and pricing are increasingly tied to Asian policy and capacity decisions, not just US cloud roadmaps.
Applied AIGSMA Ltd. CEO Hoffman on Mobile Technology
When the GSMA CEO says China is well positioned to shape agentic AI in communications, he’s flagging where the standards and default behaviors may be written. If your product depends on telco rails, track Chinese ecosystem moves as closely as US ones—interoperability and compliance will not be one-sided.