Anthropic Uses Explicit User Feedback to Train Models
October 27, 2026
Anthropic defaults to not training on inputs or outputs from commercial products like Claude API or Claude for Work. However, providing thumbs up/down feedback or reporting bugs allows the company to use those specific conversations and settings for model training.
HOW THIS AFFECTS YOU
●
builderBe aware that explicit feedback buttons can opt your users' data into training pipelines.
●
policyEnsure your data privacy compliance accounts for feedback-driven data ingestion.