Cybersecurity for Artificial Intelligence

The goal of this workshop is to discuss and debate on the cybersecurity challenges related to Artificial Intelligence (AI). Whereas undoubtedly beneficial (AI brings numerous benefits to everyday life), one should not sidestep the fact that AI and its application on automated decision making – especially in safety critical deployments such as in autonomous vehicles - might open new avenues in manipulation and attack methods, while creating new privacy challenges. Cybersecurity is therefore of paramount importance for the reliable and trustworthy deployment of AI. The much needed call for ethics in AI relies on solid cybersecurity, since it is cybersecurity that will guarantee and assure the proper implementation of said ethics. Moreover, the need to secure AI so as to be able to trust it, is also prominent in the recently published EC White Paper on AI. In this paper, risk-based approaches are promoted in particular for high criticality sectors.