As Project Manager of ENSURE-6G, it was a real privilege to contribute to Event 6 of our project series — this one focused on Privacy Preserving Machine Learning (PPML) — hosted by Telefónica Innovación Digital, Barcelona, and co-organized by University College Dublin, Ireland. The timing couldn’t have been better: the event took place during the latter part of my two-month secondment at Telefónica Innovación Digital, which meant I got to experience the organisation and delivery of the event from the inside, working alongside the TID team on the ground in Barcelona rather than coordinating it remotely.
Spread across two days, the event brought together researchers from across our consortium to share recent work on federated and privacy-preserving learning — a topic that sits right at the heart of what ENSURE-6G is trying to achieve: building 6G networks that are secure, trustworthy, and respectful of data privacy by design.
Day 1 — 7th September 2026
The first day opened with Tijana Markovic (UPORT) presenting on Decentralized Federated Random Forests for Network Intrusion Detection, followed by Thanh Linh Nguyen (TCD) on designing sustainable incentives for privacy-preserving AI in her talk, Data Stays, Intelligence Unites.
After a short break, Mohammed Elbamby (TID) spoke about personalizing distributed learning for future network applications, and Chamara Sandeepa (UCD) closed out the technical sessions with a talk on Explainable AI for Secure Federated Learning: From Poisoning Detection to Privacy Attacks. The day wrapped up with TID project introductions, giving the group a broader sense of the research happening across the host team.
Day 2 — 9th September 2026
Day two kept up the momentum. Orhan Ermis (LIST) opened with a talk on building trustworthy, secure, and privacy-preserving federated learning architectures for real-world applications, followed by Marcela Tuler de Oliveira (TUD) on trustworthy data systems across organisational boundaries — from access control to accountable data use.
Following the break, Filippo Vannella (TID) presented on knowledge distillation-driven federated learning as a service for resource-constrained edge intelligence, and Carlos Segura Perales (TID) gave a talk on P2P learning. The event closed with Bartlomiej Siniarski (UCD) introducing the Shield-6G project, offering a glimpse into complementary work happening alongside ENSURE-6G.
Reflections
Helping bring this event together as part of my secondment gave me a much richer appreciation for how much coordination goes into connecting a distributed, multi-institution consortium like ours — and how valuable it is to have that coordination happen in person, even briefly. Beyond the logistics, the breadth of talks was a great reminder of just how many angles there are to the privacy-preserving ML problem: from federated architectures and incentive design, to explainability, poisoning detection, and edge-constrained learning.
It was also a nice opportunity to connect PPML with the broader security and privacy goals of ENSURE-6G, and to see where future collaboration across these threads might lead.
Looking forward to the next event in the series.
Charuka Moremada is a PhD researcher and Project Manager of ENSURE-6G.
