Protegrity, long known for its leadership in data-centric security and for pioneering vaultless tokenization, is now extending its core philosophy into the AI era—taking aim at one of the industry’s most pressing challenges: securing data as it flows through foundation models and inference systems. With the launch of its AI Team Edition, the company is signaling a shift from traditional data protection toward a model in which security is embedded directly into AI workflows, protecting not just sensitive fields like PII, PCI, and PHI, but also the knowledge those data sets generate.
The Menlo Park–based company has unveiled the Protegrity AI Team Edition, a new security platform designed to protect data and knowledge across AI systems while addressing the growing risks posed by foundation models and evolving cyber threats.
As organizations accelerate AI adoption, Protegrity argues that data and knowledge have become critical survival factors. Improper handling or protection, the company warns, can leave enterprises vulnerable to theft, manipulation, and competitive disadvantage.
“Knowledge is what forms when facts and transactions meet context,” said Michael Howard, CEO of Protegrity. “It is inferential and lives between systems. Protegrity AI Team Edition acts as a new kind of data and knowledge firewall, enabling organizations to achieve meaningful AI outcomes securely.”
At the core of the platform is semantic-preserving encryption, a technology designed to protect not just raw data, but also the relationships and meaning that power AI models, knowledge graphs, and workflows. This enables organizations to maintain data utility while minimizing exposure risk.
The AI Team Edition aims to bridge the gap between innovation and security by embedding protection directly into AI architectures. Its capabilities extend security from data to knowledge by covering not only structured information but also the operational, behavioral, and contextual inputs used in AI systems. It enables organizations to automatically translate natural-language prompts into enforceable policies, reducing manual effort and accelerating deployment. Controls are applied end-to-end across data pipelines, analytics engines, and inference workflows, dynamically adapting based on context and user roles. Built on a Kubernetes-based architecture with CI/CD integration, the platform supports secure inference across environments with rapid scalability and updates. Additionally, it provides enterprise-wide attestation, ensuring that all interactions are validated, logged, and auditable across systems, while its pricing model is designed to make enterprise-grade AI security accessible at a fraction of traditional infrastructure costs.
Industry analysts highlight the broader organizational implications of AI security. “Securing AI has become the responsibility of the entire organization, not just security or IT,” said Grace Trinidad, Research Director for Future of Trust at IDC. “Protegrity is enabling AI security in a way that aligns with how modern enterprises operate.”
With availability starting immediately, Protegrity AI Team Edition enters a market increasingly focused on safeguarding AI systems against emerging threats—while ensuring that data remains both usable and secure.
Join our LinkedIn group Information Security Community!
