Jozu's 'Awardable' DoD Status Signals New Era for AI Battlefield Security

NEW YORK, NY – April 29, 2026 – AI security firm Jozu has achieved “Awardable” status in the Department of Defense’s (DoD) Platform One (P1) Solutions Marketplace, a move that significantly accelerates the path for its advanced AI integrity tools to reach critical defense applications. The designation validates Jozu’s on-premises technology for securing artificial intelligence models and provides a streamlined procurement channel for government agencies grappling with the immense security challenges of deploying AI in high-stakes environments.

Jozu, which specializes in creating verifiable and secure machine learning (ML) deployments, was recognized for its ability to protect AI assets from supply chain vulnerabilities through runtime, even in completely disconnected or “air-gapped” settings. The P1 Solutions Marketplace acts as a digital repository of pre-vetted, competitively assessed technologies, allowing DoD units to bypass lengthy acquisition cycles and rapidly integrate mission-critical solutions.

“The DoD operates in environments where a compromised model or rogue agent isn't just a security incident, it's a mission risk,” said Brad Micklea, CEO and cofounder at Jozu, in a statement. “Being recognized as Awardable by the P1 Solutions Marketplace validates what we've built and makes it easy for government agencies to acquire.”

Streamlining Innovation for National Defense

The Platform One Solutions Marketplace represents a fundamental shift in the DoD's approach to technology acquisition. Faced with a rapidly evolving global threat landscape, the department has prioritized speed and agility, aiming to get cutting-edge tools from the private sector into the hands of warfighters faster. The Marketplace cuts through traditional red tape by presenting solutions that have already undergone rigorous competitive evaluation based on criteria including DoD impact, technical differentiation, and mission alignment.

For a company like Jozu, achieving “Awardable” status is more than a seal of approval; it is a direct conduit to its target government customer. This pre-vetted designation satisfies many of the complex federal acquisition regulations upfront, transforming a procurement process that can often take years into one that can be completed in a fraction of the time. This efficiency is crucial for both the government, which needs to keep pace with technological advancements, and for innovative companies that can be stymied by bureaucratic friction.

The solutions featured in the marketplace are presented via five-minute pitch videos, accessible only to government personnel. Jozu’s video, titled The Jozu AI Assurance Platform–Secure Deployment for Air-Gapped AI/ML, reportedly demonstrates a real-world use case of its tamper-proof packaging and multi-layered security scanning, a key factor in its selection among a competitive field of applicants.

The Critical Need for AI Integrity

As the DoD increasingly integrates AI into everything from logistics and predictive maintenance to intelligence analysis and autonomous systems, the integrity of these AI models has become a paramount national security concern. The very data and algorithms that grant a strategic advantage are also high-value targets for adversaries. Threats range from data poisoning, where training data is maliciously corrupted to produce flawed outputs, to adversarial manipulation, where subtle, hard-to-detect inputs can trick an AI system into making catastrophic errors.

The DoD has established its own ethical AI principles—Responsible, Equitable, Traceable, Reliable, and Governable—to guide development and deployment. Jozu’s platform directly addresses the principles of traceability and reliability. By ensuring that an AI model has not been tampered with from its creation through its deployment, it provides a verifiable chain of custody that is essential for building trust in AI-driven systems.

This is particularly vital in what the company terms DDIL—Disconnected, Disadvantaged, Intermittent, and Limited bandwidth—environments. For a soldier in the field or an analyst on a submarine, an AI tool that relies on a stable cloud connection is useless. More dangerously, an AI model that has been secretly compromised could provide false intelligence or make a disastrous targeting recommendation. Jozu’s technology is engineered to function entirely on-premises, with no external dependencies, providing a self-contained and verifiable ecosystem for AI operations.

“Our AI security and provenance solution works anywhere - even in DDIL or air-gapped environments,” Micklea noted. “We're proud to bring that capability to government customers who need it most.”

An Open-Source Backbone for Verifiable AI

Underpinning Jozu’s security claims is its foundation on CNCF KitOps, an open-source standard for packaging and distributing AI projects. Governed by the Cloud Native Computing Foundation (CNCF), which also stewards technologies like Kubernetes, KitOps brings the proven principles of modern software development to the often-chaotic world of machine learning.

KitOps functions like a “Docker for ML,” bundling all components of an AI project—including model weights, datasets, code, and configurations—into a single, immutable package called a “ModelKit.” This standardization is the key to Jozu’s security promise. Each component within a ModelKit is cryptographically hashed. Any modification, whether accidental or malicious, would alter the hash, instantly revealing the tampering. Furthermore, these ModelKits can be digitally signed, allowing a CI/CD pipeline to automatically reject any unsigned or compromised AI asset before it ever reaches a production environment.

By building on an open standard, Jozu offers a level of transparency and interoperability that is often missing from proprietary, black-box AI solutions. This approach allows its platform to integrate with existing DevOps tools and container registries that the DoD already uses, lowering the barrier to adoption and leveraging existing security infrastructure. It provides a complete audit trail and provenance tracking, ensuring that every AI asset is fully traceable from its source—a critical requirement for meeting the DoD’s stringent accountability standards and building trustworthy systems for the modern battlefield.