- Protect AI and Leidos collaborate to enhance AI security in U.S. government systems.
- Partnership aims to safeguard AI across various federal agencies against emerging threats.
- Comprehensive AI security suite tackles adversarial manipulation, model drift, and more.
Protect AI, a leader in enterprise AI security solutions, has announced a strategic collaboration with Leidos (LDOS, Financial) to fortify AI systems utilized by U.S. government agencies. This partnership is aimed at delivering robust AI security capabilities that protect mission-critical applications from adversarial threats and vulnerabilities. By integrating Protect AI's platform with Leidos' expertise in secure digital transformation, the collaboration addresses the growing risks posed by agentic AI models, which can autonomously make decisions without human intervention.
Steve Hull, Digital Modernization Sector president at Leidos, emphasized the importance of this collaboration, stating that the federal government is rapidly scaling its use of AI to support national security and critical infrastructure. The partnership aims to ensure that AI systems remain safe, secure, and compliant with federal standards, thus enabling secure AI innovation at scale.
Protect AI’s platform offers a range of security tools, including Guardian, Recon, and Layer, to provide end-to-end AI supply chain protection. These tools are designed to detect and mitigate vulnerabilities in machine learning models and generative AI workflows, offering capabilities such as zero-trust security scanning, automated adversarial attack simulations, and real-time threat detection.
This collaboration aims to deliver full lifecycle security capabilities across national security, defense, intelligence, healthcare, and civil agencies, allowing for reliable and secure AI adoption. Leidos and Protect AI are committed to ensuring that government AI systems can be trusted and compliant with standards like NIST, OWASP, and MITRE, thus future-proofing the safety and security of modern software systems.
