← The AI Systems Map
L0 · LabsStandard / protocol⬦ ChokepointGlobal

Training Data & Web Corpus

The raw material of models

High-quality training data—licensed (Reddit, news, Shutterstock), scraped web (Common Crawl), and increasingly synthetic—is the true feedstock of frontier models and is now legally and commercially contested via copyright litigation and licensing deals. Data access and provenance are a genuine, non-hardware chokepoint on model capability.

How it fits the stack

Training Data & Web Corpus with what it depends on (above) and what it feeds (below). The figure renders as a crawlable diagram and upgrades to an interactive 3D graph as it scrolls into view.

depends onsourcessuppliesdepends ondepends onsuppliesusessuppliesdepends ondepends oncompetes withTraining Data & WebCorpusLabsCommon CrawlchokepointContent Licensing &Copyright RegimechokepointAgriculture AI(AgTech)AnthropicEpoch AIGPT-5 / o-seriesOpenAIWaymo / autonomousdrivingSynthetic data &distillation
Training Data & Web CorpusDepends on ↑Feeds ↓Related

Training Data & Web Corpus in the AI stack. Training Data & Web Corpus with its immediate upstream dependencies (top) and downstream dependents (bottom) in the AI value chain. Hover a node in 3D, or read the full relationships below.

Graph data (text) — 10 entities, 11 relationships
  • Training Data & Web Corpusdepends onCommon Crawl
  • Training Data & Web CorpussourcesCommon Crawl
  • Training Data & Web CorpussuppliesCommon Crawl
  • Training Data & Web Corpusdepends onContent Licensing & Copyright Regime
  • Agriculture AI (AgTech)depends onTraining Data & Web Corpus
  • AnthropicsuppliesTraining Data & Web Corpus
  • Epoch AIusesTraining Data & Web Corpus
  • GPT-5 / o-seriessuppliesTraining Data & Web Corpus
  • OpenAIdepends onTraining Data & Web Corpus
  • Waymo / autonomous drivingdepends onTraining Data & Web Corpus
  • Training Data & Web Corpuscompetes withSynthetic data & distillation