Integrated Cyber Solutions Updates White Paper on VEIL Privacy Technology

Integrated Cyber Solutions announced in a press release an updated white paper on VEIL, its privacy-preserving machine learning product. The paper reports that VEIL can compress sensitive data by between 95 and 99.96 percent while maintaining or improving predictive performance compared to models trained on raw data.

The research, authored by Jeremy J. Samuelson, Executive Vice President of Artificial Intelligence and Innovation, evaluated VEIL across multiple supervised learning tasks in healthcare, financial services, and enterprise-scale environments. It compared VEIL with differential privacy and homomorphic encryption methods, finding that VEIL outperformed these approaches in reconstruction and attribute inference tests.

The paper notes that VEIL reduces exposure of sensitive data before it enters the machine learning pipeline, rather than protecting it after ingestion. Independent academic endorsement was provided by Dr. Mohammad Tayebi of Simon Fraser University, who confirmed no affiliation or compensation from the company.

Integrated Cyber Solutions also stated that the compression achieved by VEIL could improve enterprise AI infrastructure efficiency by reducing storage, transfer, and computational requirements in certain deployment scenarios.