Protegrity VP Diego Loureda says sensitive data fed into AI agents is nearly impossible to remove once it is on a model or vector store. He argues that security teams must shift from being the ‘department of no’ to enablers of AI value. The company won the Cloudera Emerging Technology Partner of the Year award at the Cloudera EVOLVE26 event.
Protegrity’s tokenization method is designed to be quantum-ready because it relies on data replacement rather than mathematical algorithms. Enterprise Management Associates found 83% of AI leaders report project delays caused by traditional security governance. Additionally, 82% of leaders push AI agents into production in a diminished state due to stripped-down tools and data.
The firm’s technology addresses the challenge of data permanence in AI systems. Security leaders must adapt their governance frameworks to support the deployment of full-featured AI agents without sacrificing safety. This shift is critical for organizations looking to maximize the value of their artificial intelligence investments.
