Opaque Systems
Opaque Systems builds confidential AI infrastructure that lets enterprises run AI models on sensitive data using hardware-enforced privacy guarantees and verifiable proof, so organizations can adopt AI on regulated or proprietary data without exposing it to the model provider or infrastructure operator.
Visit Website ↗ + Add to CompareOverview
Opaque Systems provides what it calls a “trust layer for enterprise AI” — infrastructure that lets organizations run AI workloads on sensitive data while cryptographically and hardware-enforcing that the data itself, and the AI system’s outputs, remain confidential and governable even from the infrastructure provider. This targets a specific enterprise adoption blocker: regulated organizations that can’t risk sensitive data leaving their control to use cloud AI models.
Founded in October 2023 by a team of academics from UC Berkeley’s RISELab — including Ion Stoica, Raluca Ada Popa, and Wenting Zheng — Opaque is based in New York City and counts Intel among its investors, alongside CEO Aaron Fulkerson and Chief Platform Officer Imran Siddique, formerly of Microsoft Azure.
Innovation Matrix Assessment
Founded in late 2023 by RISELab academics with deep confidential-computing research backgrounds, Opaque has moved from research lineage to a commercial platform with named enterprise case studies within roughly two years.
Enabling AI adoption on data that would otherwise be off-limits under governance policy directly unblocks a real operational constraint many regulated CISOs currently face.
Intel's investment and case studies referencing ServiceNow and Accenture indicate credible early enterprise interest, though specific funding amounts and customer counts are not disclosed.
Hardware-enforced, cryptographically verifiable confidential computing for AI workloads is a fundamentally different trust model than policy-based data governance, addressing a root-cause adoption blocker rather than a symptom.
The underlying confidential computing techniques stem from established academic research at UC Berkeley's RISELab, giving reasonable technical credibility, though no independent third-party audit of the commercial platform specifically was found.
As AI adoption on regulated and proprietary data accelerates, confidential computing infrastructure addressing the resulting data-exposure risk is likely to grow in strategic importance.
Why CISOs Should Care
Regulated industries and organizations handling proprietary data often can't use cloud AI models at all under current governance policies; confidential computing infrastructure gives CISOs a way to unlock AI adoption on sensitive data without violating data-residency or confidentiality obligations.
What Makes It Different
Opaque's hardware-enforced confidential computing approach provides cryptographic, verifiable guarantees about data handling rather than relying on contractual or policy-based assurances from AI model providers, a meaningfully stronger trust model for regulated use cases.
The Matrix Verdict
60/100 — INCREMENTAL INNOVATOR
A technically deep, academically grounded confidential AI infrastructure vendor with credible founders and Intel backing; scores reflect strong architectural relevance to enterprise AI adoption, tempered by its early operating stage.
Editorial Note: Claims vs. Verified Findings
Founding date, founder backgrounds, and Intel's investor relationship are independently confirmed via the company's own site; specific funding amount and customer count are not disclosed.
Sources
Alternatives to Opaque Systems
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
Early-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Alice (formerly ActiveFence)
Israeli AI security company (rebranded from ActiveFence in January 2026) offering a lifecycle platform to test, guard, and…
Reco
Reco secures the "agentic ecosystem" — mapping what AI agents can access across SaaS and enterprise apps, detecting…