NexQloud Unveils Sealed, an Enforced and Verifiable AI Governance Platform Now Under U.S. Navy Evaluation for the E-2D Hawkeye

Sealed prevents out of policy AI actions in hardware and proves it did, a preventive alternative to governance tools that only observe and log. Built on four U.S. patent applications, with U.S. Air Force evaluation interest.

PALO ALTO, Calif., July 14, 2026 (GLOBE NEWSWIRE) -- NexQloud Technologies, Inc., a distributed cloud computing company, today introduced Sealed, an AI governance platform that enforces policy inside hardware and produces cryptographic proof that it did. Where most AI governance observes and logs what a model or agent did, Sealed prevents the out of policy action from happening at all, then issues a receipt anyone can verify. Sealed is built on four U.S. patent applications. The U.S. Department of the Navy is evaluating a Sealed based application, ELLMENT, for decision support aboard the E-2D Hawkeye, and the Office of the Chief Technology Officer at Air Force International Affairs (SAF/IA) has issued a Letter of Support expressing evaluation interest in the same runtime.

Two Moats: Prevention and Proof

Two properties separate Sealed from the AI governance category, and it is the only platform that delivers both:

Enforced, not observed. Sealed’s controls run inside a hardware Trusted Execution Environment. An action outside policy, an agent exceeding its scope, sensitive data leaving unredacted, or a write to an unauthorized target, is blocked before it happens rather than flagged afterward. That makes Sealed a preventive control, where nearly every governance product is a detective one.

Proven, not trusted. Every enforced decision produces a tamperproof receipt on a public log. The customer’s own security team verifies each receipt against the chip vendor and the public record. NexQloud is never in the trust path.

What Sealed Does

Sealed governs the AI an organization already uses, including closed commercial models, through gates enforced in hardware:

Capability gate. An agent receives a signed scope of read, act, and write permissions. Anything outside that scope is blocked.

Egress gate. Data leaving the enclave is minimized and redacted, with a manifest recording exactly what left and to which model.

Ingress gate. Every reply is checked for identity leakage, prompt injection, and data exfiltration before it is accepted.

Write back gate. Results are written only to permitted targets.

A policy gate routes work by data sensitivity. The most sensitive workloads stay on a U.S. origin open weight model sealed inside the enclave, where no operator, including NexQloud, can see the data. Less sensitive work can still use a frontier closed model such as OpenAI or Anthropic, but only through the egress gate, which minimizes and redacts what leaves and logs exactly what went where, with the ingress gate checking the reply before it is accepted. When data is retired, Sealed issues a verifiable erasure receipt proving it is cryptographically unrecoverable, and every gate decision is recorded in the same receipt, so the policies an organization sets are provably the policies that ran.

One Architecture, Two DoD Use Cases

Sealed is the confidential compute architecture at the core of NexQloud’s work, and it already underlies two very different applications. ELLMENT (E-2D Large Language Model Entity), the company’s U.S. Navy submission, embodies the Sealed architecture in a standalone AI assistant that gives E-2D Hawkeye flight officers traceable, source attributed answers even when cut off from external networks, submitted under the Navy Small Business Innovation Research (SBIR) program to PMA-231, the program office for the E-2D. Qlarity, NexQloud’s cloud cost intelligence layer, uses the same Sealed architecture where sovereignty and governance require it and was recently assessed Awardable by the Department of Defense Chief Digital and Artificial Intelligence Office (CDAO) Tradewinds Solutions Marketplace. The Navy submission is under competitive evaluation and is not an award, and the SAF/IA letter reflects evaluation interest, not a U.S. Government endorsement.

“Most AI governance tells you what went wrong after it already went wrong,” said Mark Hannah, Founder & Inventor of NexQloud. “Sealed stops the out of policy action in hardware before it happens, and it hands you proof you can check yourself, without having to trust us. For regulated and defense missions, that is the difference between a promise and a guarantee.”

“What makes Sealed valuable to the Navy is exactly what makes it valuable to a hospital, a bank, or an insurer,” said Robert Barbieri, Chief Executive Officer of NexQloud. “Every regulated industry wants to put AI to work on its most sensitive data and cannot risk sending it to someone else’s cloud. Sealed enforces the rules automatically and gives auditors proof instead of promises, turning a compliance blocker into an operational advantage.”

Independently Verifiable

NexQloud’s claims are open to inspection. Its Distributed Cloud Platform (DCP) is live, offering Kubernetes, virtual machines, storage, managed databases, private networking, and a container registry. The company holds an unqualified SOC 2 Type 2 attestation with a live Vanta Trust Center, a public SEC Regulation CF filing, and a live fleet of 1,800+ ISV certified commercial workstations at livefeed.nexqloud.net.

About NexQloud

NexQloud Technologies, Inc. builds secure, verifiable cloud infrastructure for sensitive and regulated workloads. Sealed, its AI governance platform, enforces policy inside hardware and proves it, so organizations can put AI to work on sensitive data with control that is prevented rather than merely monitored, and proof they can verify themselves. Qlarity, its cloud cost intelligence layer, gives one view of spend across every cloud, a single efficiency score, and savings the AI identifies and an operator approves and can reverse. NexQloud is on track for CMMC Level 2, is pursuing FedRAMP® Moderate authorization, and is a U.S. owned, FOCI clear small business. For more information, visit www.nexqloud.io.

Forward Looking Statements

This press release contains forward looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. The referenced patent applications are provisional and may not result in issued patents. Participation in the SBIR program is a competitive, multi stage evaluation and does not constitute a contract, an award, or any assurance of selection, and a Letter of Support reflects evaluation interest rather than endorsement or commitment by the U.S. Government. Statements regarding CMMC Level 2 and FedRAMP® Moderate authorization reflect current plans and timing that may change. NexQloud undertakes no obligation to update any forward looking statement.

Media Contact

Robert Barbieri, Chief Executive Officer
Email: b.barbieri@nexqloud.io
Phone: +1 669 241 0916
www.nexqloud.io

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.