
BIG DATA NEEDS TRUSTWORTHY TECHNOLOGIES
How innovative technologies will help to realize the potentials of big data and improve collaboration. Stefanie Lindstaedt provides insights into the future with trustworthy AI technologies.
Data is being generated 24/7. Properly linked, it can be a source of various new ideas and business models. With the help of big data and AI, production and supply chains can become more flexible and regional. This allows to create an economic structure that is resilient to global catastrophes, e.g., the current pandemic.
To establish such a resilient circular economy connection and optimization of entire production systems is required and large amounts of data from many sources have to be collected, shared and evaluated. Experts and companies from different fields must work together collaboratively. This requires appropriately networked computer and data analysis environments, as well as the ability to share the results while maintaining data protection.
Europe takes the lead in data protection
In the public opinion, a contradiction between big data and privacy exists: the more data is used, the more likely it is that privacy will be jeopardized.
Therefore, sharing and analyzing sensitive data requires clear ethical and legal frameworks. Europe is taking a unique path and is prepared to lead the way into a data protective future. The first step towards this goal was introducing the GDPR to prevent the misuse of data. It is a double-edged sword. While the GDPR fulfils the task of protecting data, it also reduces the willingness to use data for solving economic and social problems. Therefore, we need new solutions that reconcile the secure and verifiable evaluation of sensitive data with privacy protection.
Next level trustworthy technologies
The lack of confidence in data security is an obstacle to use big data and AI to establish new business models. These privacy challenges can be tackled using innovative cryptographic methods, such as homomorphic encryption, that enable a secure collaboration among stakeholders by allowing evaluations of encrypted data. This way, sensitive information can be shared for calculations or train AI models while keeping its content private. Such AI-powered solutions have the potential to enhance the ecosystems massively.
In the health sector, for example, sensitive data can be used to improve diagnosis and therapy. In production supply chains could be optimized across suppliers. Another AI-supported solution is federated learning. Here, an AI model is trained and evaluated using different devices or clients. The model parameters are exchanged exclusively via the network, while the training or evaluation data is stored locally. This creates many new possibilities and new business models. For example, delivery ecosystems in production across suppliers could be optimized without putting private data at risk.
Trustworthy AI and ethics shape Europe’s future
In EUH4D, the overall goal is to provide services with “the highest level of privacy” to SMEs and start-ups as soon as possible. By providing easy, cross-border access to datasets and facilitating data sharing new innovative data-driven solutions and business models can be established and the vision of “Collaborative European Data Spaces” be implemented.