Safeguarding ethical values when storing and processing data and applying algorithms is crucial to maintain trust in our systems and our organisation.

Personal data governance

 

Our data is omnipresent in our lives and in society. As citizens, we are becoming more aware of the use of data by third parties and its importance. Knowledge centres, universities, governments, and companies need data to be able to develop new technologies, do research into evolutions, diseases, etc.

Still, we should always retain control over our personal data. Personal data governance is a big part of many VITO initiatives, such as We Are and the 8 effectiveness principles for ‘caring technology’.

Trustworthy AI

 

The upcoming EU Artificial Intelligence Act will forever change the way AI is created. Pursuing trustworthy AI by design is the focus of a new line of research at VITO. The goal is to improve the technology, so guarantees can be made on the AI performance outside of the lab.

These guarantees are crucial to get AI certified as a medical device, where it can provide a lot of added value. New technology is developed to continually monitor model performance after deployment, and to better predict real-world model performance taking model robustness into account.

The international Vivaldy PENTA project is the one of the first projects to put unlocked, trustworthy AI at the centre. Vivaldy addresses re-verification and re-validation by analysing field obtained user/usage data, change-based impact analysis, and integrated virtual and physical testing.