Digital data is indispensable for any AI-based work. In order to be of true value such data must be trustworthy and available at anytime, from anywhere. Data is in fact the fuel needed for much of modern AI, becoming increasingly indispensable in today’s science.
Prior to the analysis of the data, collecting and managing useful data is often a major challenge. Things can get particularly difficult when personal data, intellectual property and privacy issues are involved. A further challenge can be combing different data sets, but subject to ethical and legal constraints.
Here at EAISI, our goal is to develop technologies and protocols that provide trustworthy, anytime, anywhere data. This involves creating new technologies for the safe anonymization or pseudonymization of existing data, and the synthetization of new data when sources are not available. These methods must also handle this automatically and on-the-fly because of the demands for quick responsiveness in health, industry and mobility. These must resolve the challenges of restricted or biased data sets, while ensuring the privacy of the subjects concerned.