Summary
Anthropic has introduced Enterprise Frontier Safeguards (EFS) to monitor AI usage for safety without compromising enterprise customer data, storing activity data within customer-controlled cloud infrastructure. While prompts and outputs won't be stored, some data will be analyzed due to concerns about AI misuse for cyberattacks, with flagged issues sent to customers for review.