Larry Ellison-backed argument says training data may be AI’s last moat

chrisgrayson · x · 2026-07-22

The post argues that training data may become the last real moat for AI models, especially when a prominent industry voice says the same thing.

It quotes Larry Ellison in support of the idea that the data used to train models will matter more and more, then points to examples like LexisNexis owner RELX, which has a $55B market cap.

The broader point: as models commoditize, proprietary content and data access may become the strategic advantage startups and incumbents fight over.

Related event: Training Data Emerges as the Ultimate AI Moat(2 posts)→

Original post →

More from Companies & People

Companies & People channel →