← All systems
Training-Data Pipeline
Web content is scraped into a dataset and trained into a model
Architecture introduced 20 Feb 2023
Big models learn from huge piles of web content, scraped automatically. But the web changes — and an attacker who controls even a tiny slice of what gets scraped can quietly teach the model something wrong.
InstructionsDataActionsControl / decisionFeedback / logs
👆 Click any component in the diagram to inspect its risks & defensesFollow a request · step 1 of 3
← / → keys
A crawler grabs content from millions of web addresses.