Anthropic Licensed Confidential Computing Tech From Opaque
The partnership integrates secure data processing tools into Anthropic's enterprise AI offerings.
Updated on Oct. 14, 2026 in Artificial Intelligence
Live Poll
Do you trust AI companies to adequately protect your sensitive data when using new security technologies?
Anthropic has licensed confidential computing technology from Opaque Systems and hired Opaque CEO Aaron Fulkerson. The agreement, announced October 13, 2026, aims to bolster secure data handling for enterprise AI workloads.
Why it matters
This move enhances Anthropic's capability to process sensitive enterprise data by leveraging hardware-enforced security protections. The addition of Opaque’s expertise signals a push to satisfy strict security and compliance requirements for corporate clients.
Opaque Systems' platform enables the execution of workloads on sensitive data through hardware-enforced protections. The system produces cryptographic evidence that verifies how specific data was processed and which internal policies were applied.
The players
Anthropic
An AI research and deployment company that develops large language models and focuses on enterprise safety.
Opaque Systems
A startup originating from the UC Berkeley RISELab that specializes in confidential computing and secure data processing.
Aaron Fulkerson
The former CEO of Opaque Systems who is joining Anthropic following the licensing of his company's technology.
The details
The technology leverages confidential computing—a method of isolating data in secure hardware enclaves while it is being processed. By generating cryptographic evidence, or verifiable mathematical proofs, the system ensures that AI models interact with data only under specified, audited constraints. This mechanism allows organizations to maintain data privacy even when offloading compute tasks to third-party AI platforms.
Timeline
October 13, 2026: Opaque Systems announced the license agreement.
The Tech Race
This integration follows a broader industry trend of moving sensitive computation into secure, verifiable enclaves. It places Anthropic in a stronger competitive position to offer enterprise-grade AI while addressing the data privacy concerns that often stall corporate adoption.
Enterprise customers will likely see new data security features integrated into Anthropic's AI products in future releases. Exact timelines for when these security improvements will be available for production workloads remain unannounced.
The takeaway
Anthropic is prioritizing hardware-backed security as a competitive differentiator in the enterprise AI market. Watch for future product roadmap updates regarding the integration of cryptographic verification features for enterprise users.
Further reading
For broader context on how providers are securing model inputs, explore the Artificial Intelligence section.
Live Poll
Do you trust AI companies to adequately protect your sensitive data when using new security technologies?
