|Business||Period||Project Coordinator||Funding Scheme|
|Industry||July 2022 -December 2025||FORTH Foundation for Research and Technology - Hellas||Horizon Europe|
Ensuring privacy and trustworthiness for data owners and providers presents a myriad of complex challenges.
Data breaches remain a prominent threat, with Europe as the most affected in Q1 2023 (17.5 million), followed by Asia and North America.
Data encryption and anonymization are critical to safeguard data from unauthorized access, but striking a balance between strong encryption and data usability can be tricky. Secure data sharing between multiple parties is necessary for collaboration while preserving privacy, but it often demands computationally expensive techniques like secure multi-party computation. This led data owners to adopt data minimization practices for reducing privacy risks, but this approach must also align with business needs.
Additionally, obtaining user consent and providing transparent information about data processing practices are essential for building trust, but designing user-friendly consent mechanisms can be challenging.
From an economic perspective, the cost of compliance with robust privacy measures and data protection practices poses significant challenges, particularly for smaller companies and startups. Liability and legal concerns add economic risks, making it vital for data owners and providers to take proactive steps in managing potential legal consequences. Monetizing data ethically while respecting privacy is another economic challenge. Striking a balance between generating value from data and ensuring data privacy is a delicate task that necessitates careful consideration involving the development of economic models that provide fair incentives and compensate for potential risks, encouraging for example data sharing for research and collaboration.
Standardizing privacy regulations and technical measures across regions can help reduce compliance challenges and related costs. Moreover, third-party services are often relied upon for data processing, storage, and analysis. Ensuring these service providers adhere to strict privacy standards and data protection practices is crucial for maintaining trust in the data ecosystem.
Addressing these multifaceted challenges requires a comprehensive approach, involving strong technical measures, clear regulations, effective enforcement, and ongoing collaboration between stakeholders. Only through collective efforts can we establish a data landscape that prioritizes privacy and trustworthiness for all involved parties.
In this challenging scenario, RINA is committed to support data scientists and AI developers and bring benefit to European data space stakeholders.
Using self-sovereign technologies and with state-of-the-art homomorphic encryption, we participate in the TRUSTEE project aiming to bring a green, secure, trustworthy, and privacy-aware framework able to aggregate and make available various interdisciplinary data repositories, such as Healthcare, Education, Energy, Space, Automotive, Cross-border Data Exchange, as well as consider other European data federation spaces and trans-national initiatives, such as Gaia-X and EOSC.
In particular, we provide the Trustworthy AI Support Design Framework built on human-centric processes and a set of supporting tools to ensure lawful, ethical, and robust AI throughout the product lifecycle.
Our approach is applicable for AI software development to make AI-based systems as robust and resilient to attacks as possible through various measures:
- adherence to ethical principles based on fundamental rights (e.g., fairness and explainability)
- risk analysis considering safety, security, and privacy
- continuous monitoring and adversarial testing for AI robustness, adopting recent programming practices and innovative secure deployment methods.
The Framework considers a typical assessment and system monitoring audit workflow as step-by-step support to develop explainable and trustworthy AI applications.
Using TRUSTEE and RINA solution, AI developers and data scientists can benefit from trustworthy technologies, privacy-preserved data sharing, Cloud to Edge to IoT and Explainable AI by:
- boosting European leadership in the global data economy
- wide spreading trusted data solutions and generating social and environmental positive impact
- automate flexible testing and monitoring of AI assets
The Framework brings a remarkable added value to support AI system developers and product managers across:
- ideation, to clarify the expected outcomes and analyse competing hypotheses
- planning, to understand the interfaces, type of feedback of the AI system and gather requirements in compliance with regulatory and ethics principles
- design, to define dataset semantics and monitor the evolution of the algorithms
- build, to ensure the AI system is resilient
- test, to validate the system functionalities and security features
- deploy, to provide deployment containers security
1. FORTH Foundation for Research and Technology - Hellas (Project Coordinator) 2. Elliniko mesogeiako panepistimio 3. Panasonic automotive systems europe GMBH 4. Universitat wien 5. K3Y 6. Acceligence LTD 7. Universita Cattolica Del Sacro Cuore 8. Ericsson Nikola Tesla D.D. 9. Adrestia erevnitiki idiotiki kefalaiouxiki etaireia 10. Fundacion Tecnalia Research & Innovation 11. Rina Consulting SPA 12. Zortenet idiotiki kefalaiouxiki etaireia 13. Inqbit Innovations Srl 14. Teknologian Tutkimuskeskus VTT Oy 15. Fujitsu Technology Solutions (Luxembourg) Sa, 16.Hewlett Packard Italiana Srl 17. Hewlett-Packard Customer Delivery Services Italia Srl 18. Institutul De Stiinte Spatiale 19. Checkwatt Ab 20. Athina-Erevnitiko Kentro Kainotomias Stis Technologies Tis Pliroforias 21. Ton Epikoinonion Kai Tis Gnosis 22. Etoile Partners Ltd 23. Edinburgh Napier University