GMV’s Big Data and AI Division takes part in the 15th Cyber Commanders Forum summit

Málaga hosted the 15th edition of the Cyber Commanders Forum (CCF) on September 21-23, under the leadership of Vice Admiral Javier Roca Rivero. As part of this prestigious forum, the Andalusian Digital Agency hosted the event “Cybersecurity Strategies from Southern Europe” on the final day, with GMV playing a key role. During this meeting, leaders from the security agencies of NATO, Japan, Brazil, and Australia—counterparts of the Joint Cyber Command (JCSCM)—gained firsthand insights into the industry’s vision and innovative proposals.

Cyberspace has enabled a phenomenon of transnational terrorism with a global impact, capable of inflicting damage indiscriminately. Combating these new threats is central to the strategy of international security and defense organizations. In this context, during the “Cybersecurity Strategies from Southern Europe” event, leading companies in the sector, including GMV, showcased innovative solutions aimed at tackling these challenges.

José Carlos Barrios, project manager in the Big Data and Artificial Intelligence division of GMV’s Secure e-Solutions, emphasized how the company's focus on innovation has resulted in tools that help the Armed Forces protect their assets. Among the products presented were:

- PitIA: A solution for anomaly detection through data analysis.
- PenBot: A tool that performs automated tests to identify system vulnerabilities.
- uQuery: This solution optimizes decision-making by leveraging data to extract valuable insights and facilitate easy, intuitive access to information, using large language models (LLMs).
- uTile: Its functionality was explained in collaboration with military personnel from Algeria and Ireland, who participated as members of the federated AI network.

Barrios emphasized the importance of Privacy Enhancing Technologies (PET) in developing uTile, highlighting their advantages over the traditional approach of centralizing data and the risks of sensitive information leaks. He noted: “Accessing data under a standardized model is complex, as it is often sensitive and subject to regulations and strict sharing protocols. Limited data availability can compromise the accuracy of our machine learning (ML) models or statistical analyses.”

However, with uTile, “accessing the data is unnecessary. Through technology developed with open-source solutions, data remains at its source, preserving privacy while enabling decentralized use.” Some of the techniques implemented in uTile to protect privacy and ensure data security include Federated Learning (FL), Secure Multi-Party Computation (SMPC), Private Set Intersection (PSI), Homomorphic Encryption (HE), and Zero-Knowledge Proofs (ZKP).