Anthropic said this week it is introducing an enterprise Frontier Security framework, or EFS, aimed at a long-running issue for business users: how to support zero data retention while still detecting abuse patterns. Under the design, monitoring data is stored inside cloud infrastructure controlled by the customer rather than on Anthropic’s own servers, separating data sovereignty from the company’s safety review process. The system has not been fully rolled out yet and is scheduled for a phased launch later this fall, with early access available through an application process. Anthropic said the framework was developed with more than 100 customers across financial services, healthcare and manufacturing, along with Amazon Web Services, Google Cloud and Microsoft Azure. The company added that suspicious signals identified by the system are sent directly to customers for human review, while Anthropic continues to operate the automated abuse-detection layer. The architecture is designed to catch complex misuse spanning multiple tasks, sessions and accounts, including attacks tied to stolen enterprise credentials.
Anthropic said this week it is launching an enterprise Frontier Security framework called EFS to address a core tension for business customers: the trade-off between zero data retention, or ZDR, and abuse detection.
Under the new design, monitoring data is stored in cloud infrastructure controlled by the customer rather than on Anthropic servers. That setup separates data sovereignty from security monitoring.
Rollout timeline
The framework has not been fully deployed. Anthropic plans to release it in phases later this fall, with initial access handled through an application-based process.
Who helped build it
According to the company, EFS was built with more than 100 customers across financial services, healthcare and manufacturing, as well as Amazon Web Services, Google Cloud and Microsoft Azure.
How the system works
Anthropic said one of the main design choices was moving monitoring data storage to the customer side. When suspicious patterns are detected, signals are sent directly to the customer for human review. Anthropic will still run the automated abuse-detection system.
The framework is intended to identify complex misuse that spans multiple tasks, sessions and accounts, including attacks involving the theft of enterprise credentials.
Company statement on training data
Anthropic said it has never used enterprise data to train models without explicit permission. The company added that its earlier 30-day data retention policy was introduced solely to improve detection quality.
EFS was announced alongside Claude Fable 5.1 and Mythos 5.1 as one of three enterprise-focused concessions, according to MarkTechPost.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
