Copyright, consent, PII, bias, and freshness all live in the training set. Increasingly the questions regulators, courts, and buyers ask are 'what is in it' and 'who said you could use it'.
GlossaryPart of the BotDoors AI Safety & Compliance Glossary · 37 terms.
Ethics
Training Data
Training data is the corpus a model learns from; its provenance, licensing, and quality bound everything the model can do.