Regulation and security
Claude researches ways to reduce AI safety failures
Anthropic used Claude to research methods for reducing ten types of unwanted AI behavior.
Read publication →EDITORIAL COLLECTION
Regulation and security
News and explanations about technology regulation, privacy, and security.
3 publicationsContinuously updated
PUBLICATIONS
Explore this category
Regulation and security
Anthropic is preparing a text watermark for Claude
Future Claude models will add a statistical signal to estimate whether AI participated in a text.
Read publication →
Regulation and security
OpenAI expands data protection in its API
OpenAI offers Zero Data Retention to eligible API customers and is preparing safety controls for longer-running tasks.
Read publication →