As enterprises struggle to move AI initiatives from pilot to production due to data leakage risks, Opaque has raised $24 million in Series B funding to bridge the “trust chasm.” The round was led by Walden Catalyst, with participation from Intel Capital and the Advanced Technology Research Council, valuing the San Francisco-based company at $300 million.
The big picture: Most companies pause AI deployments when sensitive proprietary data is involved. Opaque provides a “confidential-first” infrastructure that delivers cryptographic proof that data remains private during computation, model weights are never exposed, and policies are enforced exactly as written.
How it works:
- Confidential Enclaves: Executes workloads within hardware-protected environments that isolate memory from underlying systems.
- Verifiable Governance: Extends traditional observability tools by providing runtime attestation that privacy rules were followed.
- Opaque Studio: A development environment for building AI agents that are “audit-ready” before, during, and after execution.
The catch: While Opaque promises a “trust layer,” its architecture relies heavily on specific hardware-backed protections from chipmakers like Intel and NVIDIA. This creates a high level of vendor lock-in and potential performance overhead that could slow down real-time AI agents. Furthermore, the complexity of cryptographic verification often requires a specialized skill set that many overstretched enterprise security teams lack, risking a scenario where “verifiable trust” becomes just another unmanaged layer of technical debt.
Key Details
- Funding: $24M (Series B)
- Lead: Walden Catalyst
- CEO: Aaron Fulkerson
- Sector: Cyber Security / Confidential AI
