Research testbedIn build
Federated Learning Testbed
A shared environment for training models across separate data holders without pooling their data.
fl.didlab.orgOpens soon
Research
Research that becomes practice
We study how digital systems fail, how trust can be verified, and how privacy can be protected, then turn the findings into testbeds, prototypes, learning scenarios and shared knowledge.
Research areas
Current areas of work
- AI and LLM securityModel inversion, membership inference, model extraction, data poisoning, adversarial machine learning and the security of AI-enabled applications.
- Privacy-enhancing technologiesDifferential privacy, homomorphic encryption, secure collaborative computation and privacy-preserving federated learning.
- Distributed trust and blockchainVerifiable records, attestations, provenance, blockchain-enabled traceability and decentralized identity with selective disclosure.
- Cyber-physical systems and IoTSecuring connected and intelligent devices, emerging attacks, layered defenses and verifiable coordination between systems.
- Cybersecurity educationCyber ranges, experiential learning and evidence-based assessment of practical skill.
- Knowledge and open innovationOrganizing guidance, standards, findings and lab experience into reusable resources, tools and reference implementations.
Testbeds
The platform is the testbed
The same infrastructure that runs learner labs supports research experiments: isolated networks, reproducible environments and captured evidence.
Project pages and publications will be listed here as they are released.
Collaborate with DIDLab
We work with researchers, educators and organizations on joint projects, testbeds and student research.