Anthropic has unveiled Enterprise Frontier Safeguards, a data security framework built with feedback from healthcare organizations and more than 100 other enterprise customers to detect cyberattacks and other misuse while letting hospitals keep control of their AI activity data.
EFS pairs zero data retention with automated monitoring for signs of serious misuse, including attempts to develop offensive cyber capabilities, stolen or leaked credentials, and destructive behavior by AI agents acting across multiple sessions, according to a Sept. 1 news release. Customers store activity logs in their own cloud accounts — such as Amazon S3, Azure Blob Storage or Google Cloud Storage — under their own encryption keys and audit controls, with flagged activity routed to their own security teams rather than Anthropic staff for review.
The framework directly answers a concern health system CIOs have raised with Becker’s since Anthropic began testing its cybersecurity-focused Mythos-class models this spring: that hospitals need a seat at the table as frontier AI models capable of finding and exploiting software vulnerabilities move toward wider release.
EFS will roll out in phases starting this fall and will be supported on Claude Code, Claude Enterprise, the Claude Platform, Amazon Bedrock, Google’s Agent Platform and Microsoft Foundry. Eligible customers will get zero data retention on Claude Fable 5 and Fable 5.1 in the meantime, addressing friction created since Anthropic introduced 30-day data retention with Fable 5.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
