AI is rapidly becoming the backbone of society, powering everything from healthcare to transport. Uni Systems Luxembourg leads the HORIZON cPAID project to contribute to secure AI by-design.
The rapid computing growth has driven the widespread integration of Artificial Intelligence (AI) into daily life, industries, and critical applications, where Machine Learning (ML) and Deep Learning (DL) algorithms process large volumes of data to support crucial decision-making.
Furthermore, the adoption of ML/DL technologies in place of traditional methods exposes early adopters, as insufficient protective mechanisms enable adversarial AI attacks, making AI systems increasingly attractive targets for malicious actors.
That’s where the cPAID project—Cloud-based Platform- agnostic Adversarial AI Defence Framework, steps in, with a mission to build resilience into AI itself, protecting critical systems before they’re compromised. Launched in 2024 with a lifespan of 3 years, the cPAID brings together 17 organisations across Europe with the vision of researching, designing, and developing a cloud-based platform-anostic defense framework for the holistic protection of AI applications against adversarial attacks. throughout their entire lifecycle.
Designing Resilient and Adaptive Defences
cPAID is built around well-known security strategies, such as DevPrivSecOps, AI-based anomaly detection, penetration testing, risk management, and cyber range which are reformed to the specific needs of AI systems and adversarial attacks landscape. The project is developing tools to:
- Test AI against simulated cyberattacks
- Monitor behaviour for abnormalities
- Adapt dynamically to emerging risks
It extends established practices in software development by embedding privacy and explainability into every stage of AI, providing to end users assurances about security and privacy of their AI applications.
Tested on Real World Scenarios
To ensure its solutions are practical and impactful, cPAID will be tested in real use cases scenarios.
The project includes five pilot projects, each targeting a critical domain:
- Energy: worker robots monitoring EV batteries in Estonia.
- Surveillance: 5G-enabled drones detecting wildfires in Poland.
- Health: securing remote, AI-assisted medical devices in Greece.
- Transportation: enhancing object detection for autonomous ships in Norway.
- Cybersecurity Awareness: training experts to simulate real-world threats.
Securing AI for Tomorrow
cPAID embeds AI security by design, enabling trusted innovation while protecting users and strengthening Europe’s digital sovereignty. Its initiatives guide AI’s growth to be secure, reliable, and aligned with social values.
The project funded under Grant Agreement No. 101168407 is supported by the European Cybersecurity Competence Centre. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Cybersecurity Competence Centre. Neither the European Union nor the granting authority can be held responsible for them.
