Anthropic acquires Opaque Systems' technology and key engineers
Summary
Anthropic has announced a new acqui-hire deal with Opaque Systems, acquiring the company’s technology for running AI on confidential data and hiring key engineers, including CEO Aaron Fulkerson and CTO Rishabh Poddar. This technology addresses a significant gap in standard encryption by allowing AI jobs to operate in hardware-sealed environments, thereby producing cryptographic evidence that enables organizations like banks and hospitals to audit access to their records. Previously, Anthropic had utilized Opaque's tools and published a confidential inference design with Pattern Labs in June 2025, highlighting their prior collaboration before this expanded agreement.
Analysis
Anthropic: Anthropic develops advanced AI systems with a focus on safety and reliability, including its Claude family of models. The company is licensing confidential data processing technology from Opaque Systems and hiring select engineers to strengthen its secure inference capabilities for sensitive workloads. Opaque Systems: Opaque Systems builds a platform that enables AI models to process encrypted data inside hardware-protected environments while generating verifiable cryptographic proofs of execution. It is entering a licensing and acqui-hire arrangement with Anthropic that transfers its technology and key technical leadership to the larger AI lab. Rishabh Poddar: Rishabh Poddar is CTO of Opaque Systems and has driven technical work on hardware-sealed AI execution environments. He is among the engineers being hired by Anthropic through the same transaction. Aaron Fulkerson: Aaron Fulkerson serves as CEO of Opaque Systems and has led development of its confidential AI computing platform. He is transitioning to Anthropic as part of the licensing and hiring deal announced in the news. Use Cases: The platform produces cryptographic evidence that lets organizations such as banks or hospitals audit exactly which code accessed their records and under what rules. Prior Work: Anthropic had already used Opaque’s tools and published a confidential inference design with Pattern Labs before this expanded agreement. Technology: Opaque’s approach addresses the gap where standard encryption leaves data exposed once decrypted in memory during AI processing.
Categories
techaimachine_learning