New collaboration delivers robust AI security capabilities to safeguard mission-critical government applications from adversarial threats and vulnerabilities

SEATTLE & RESTON, Va.: #AIsecurity--Protect AI, the leading provider of enterprise AI security solutions, is collaborating with Leidos (NYSE: LDOS), a leader in secure digital modernization and mission-critical technology innovation, to strengthen security for AI systems used by U.S. government agencies. The collaboration combines Leidos’ expertise in secure digital transformation with Protect AI’s platform to enhance end-to-end protection against AI threats—including those posed by next-generation agentic AI models.​

An agentic model is a type of AI system that is given the autonomy to make certain decisions and take actions without human intervention. These systems can pose risks to government agencies responsible for protecting national security, infrastructure, and other critical systems because they can make decisions that might be influenced or manipulated by outside threats. Leidos and Protect AI aim to deliver full lifecycle security capabilities across the entire AI supply chain for national security, defense, intelligence, healthcare, and civil agencies that align with government standards and enable secure AI innovation at scale.​

“As powerful and increasingly agentic AI models emerge, ensuring trust and compliance in government AI systems is more important than ever,” said Steve Hull, Digital Modernization Sector president at Leidos. “The federal government is rapidly scaling its use of AI to support national security and critical infrastructure, and our collaboration with Protect AI increases our ability to future-proof the safety and security of modern software systems.”

“AI security isn’t just about protecting critical data—it’s about securing the entire AI supply chain," said Leidos Chief AI Officer Ron Keesing. “As agentic AI models introduce new risks, our work with Protect AI aims to ensure agencies maintain control and security while also being able to move quickly. Responsible AI adoption depends on trust, and by integrating Protect AI’s capabilities, we’re strengthening defenses against adversarial threats and helping to ensure AI remains a reliable asset for mission-critical applications.”

By integrating Protect AI’s platform into Leidos’ secure digital transformation initiatives, the firms aim to provide federal agencies with critical capabilities to manage AI risks—from model selection and development to deployment and ongoing monitoring. This includes protection against issues like prompt injection, adversarial manipulation, model drift, and more.​

“Government AI systems are attractive targets for adversaries,” said Ian Swanson, CEO of Protect AI. “Our collaboration with Leidos brings full-stack AI security into government environments—ensuring every stage of the AI lifecycle is secure, auditable, and compliant with federal standards like NIST, OWASP, and MITRE.”​

The Protect AI platform provides a comprehensive suite of AI security tools designed to detect and mitigate vulnerabilities in machine learning models and generative AI workflows, that include:​

• Guardian: A zero-trust security scanner for AI/ML models, identifying backdoor threats, runtime risks, and integrity issues in open-source and proprietary models in under a minute.​

• Recon: An automated red-teaming solution for generative AI that simulates adversarial attacks, recommends security guardrails, and validates model defenses using an extensive threat library updated weekly.​

• Layer: An LLM runtime security platform that provides deep telemetry and real-time threat detection across enterprise AI applications. Layer enables both agent-based and agentless deployment options, allowing federal agencies to monitor and secure LLMs without introducing latency or operational overhead.

For more information and demo of the Protect AI platform visit www.protectai.com.

For more information about Leidos’ AI security initiatives visit Leidos.com.

About Protect AI

Protect AI addresses unique enterprise AI security concerns with a comprehensive suite of products for advanced AI scanning, robust LLM security, and GenAI red teaming that operate on one enterprise ready platform. By integrating with existing AI and security operations, Protect AI enables organizations to safeguard AI investments, enhance operational efficiency, adopt MLSecOps and adhere to global and US data security and privacy standards. Founded by AI leaders from Amazon and Oracle, Protect AI is backed by top investors, including Acrew Capital, boldstart ventures, Evolution Equity Partners, Knollwood Capital, Pelion Ventures, 01 Advisors, Samsung, StepStone Group, and Salesforce Ventures. The company is headquartered in Seattle, with offices in Berlin and Bangalore. For more information, visit our website and follow us on LinkedIn and X.

About Leidos

Leidos is an industry and technology leader serving government and commercial customers with smarter, more efficient digital and mission innovations. Headquartered in Reston, Virginia, with 48,000 global employees, Leidos reported annual revenues of approximately $16.7 billion for the fiscal year ended January 3, 2025. For more information, visit www.leidos.com.

Fonte: Business Wire