MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641064151 A) filed by Srm Institute Of Science And Technology Tiruchirappalli on May 21, 2026, for A Federated Learning For Collaborative Threat Intelligence.
Inventors include Deepith A; Eric Antony Moses; and Dr B. Kiran Kumar.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: A federated learning–based system for collaborative malware threat intelligence is disclosed, enabling secure, privacy-preserving cooperation among multiple independent organizations without requiring the exchange of sensitive raw data. The system is designed to address the growing complexity and scale of modern malware while overcoming legal, regulatory, and competitive barriers that traditionally limit cross-organizational threat intelligence sharing. In the disclosed system, each participating entity operates one or more autonomous nodes responsible for local malware analysis. These nodes execute malware samples within isolated, containerized environments to ensure system safety and reproducibility. The containerized execution environments allow dynamic and static malware analysis to be performed without exposing host infrastructure, while also ensuring consistent analysis conditions across participating nodes. Feature extraction and behavioral profiling are conducted locally, generating structured intelligence such as system call patterns, network behaviors, memory access indicators, and file system modifications. By combining federated learning with containerized malware analysis, the disclosed system enables secure, scalable, and privacy-preserving collaboration among organizations. This approach significantly enhances collective malware detection capabilities while maintaining strict control over sensitive data, thereby representing a substantial improvement over traditional centralized threat intelligence sharing models.
Disclaimer: Curated by HT Syndication.
