UNITED KINGDOM / AGILITYPR.NEWS / October 08, 2025 / New framework HFAN-Priv achieves near-perfect predictive accuracy while safeguarding employee privacy, marking a pivotal advancement in ethical AI adoption for HR.

Jay Barach, IEEE Senior Member, published author, and award-winning innovator, has announced the publication of his latest paper in IEEE, Federated Learning for Privacy-Preserving Employee Performance Analytics. The research introduces a novel federated learning framework, HFAN-Priv, designed to predict employee resignation risk and performance outcomes without exposing raw HR data.

HFAN-Priv represents a breakthrough at the intersection of artificial intelligence, data privacy, and human resources technology. By distributing model training across multiple sources of HR data without centralizing sensitive information, the framework protects employee confidentiality while delivering predictive insights of unprecedented precision.

Early results demonstrate that HFAN-Priv achieved 99.16% accuracy across key metrics, challenging the long-standing assumption that organizations must trade privacy for performance in AI-driven analytics.

“AI should empower organizations while respecting employee trust. With HFAN-Priv, we’ve proven it’s possible to deliver near-perfect predictive accuracy while guaranteeing privacy and compliance. This represents the future of ethical workforce analytics,” said Jay Barach.

The paper details several innovations that position HFAN-Priv as a benchmark for responsible AI in HR:

• Hierarchical dual attention to identify which workforce factors - such as training hours or performance scores - most influence predicted outcomes.
• Differential privacy protocols to secure updates and prevent potential data leakage between parties.
• Explainability techniques including SHAP and LIME, enabling HR professionals to understand why predictions are made, while maintaining confidentiality.
• 
  

These technical advancements mean that the framework not only delivers highly accurate predictions but also meets the rising demand for explainable, compliant, and human-centered AI systems in workforce management.

Beth Verman, CEO and President of Systems Staffing Group, Inc., praised the publication as an example of leadership in responsible innovation.

“Jay’s work demonstrates how cutting-edge AI can be harnessed to solve real-world organizational challenges without compromising ethical standards,” she said. “HFAN-Priv sets a new benchmark for the future of HR analytics. This paper marks the latest recognition in Barach’s career of contributions at the nexus of technology, privacy, and workforce innovation.”

Barach’s publication comes at a time when frameworks that reconcile innovation with evolving global regulations and organizational trust are in high demand. Thanks to its high performance with ethical safeguards, HFAN-Priv offers a viable pathway for businesses seeking to adopt AI responsibly in talent management, recruitment, and workforce planning.

For more information, visit https://thebarach.com. To read the full paper, visit https://ieeexplore.ieee.org/document/11087544.