Machine Learning has become an indispensable technology that allows us to extract insights from vast quantities of data in many industries and applications. Advances in areas such as perception, language recognition, medical diagnosis, or self-driving have been spectacular. In security, ML has been instrumental in identifying threats, attacks, and abnormal activities. However, ML algorithms have not been designed to operate in the presence of adversaries. Furthermore, securing and defending ML systems is very hard because we lack theoretical tools for developing principled ML defenses. Can ML/AI give defenders the upper ground or are we consigned to another security whack-a-mole? Panelists will discuss Federal Government’s research in ML/AI and cybersecurity and issues to drive further R&D.

Panelists:
Dr. Kenneth Calvert, Division Director, NSF/CISE
Dr. Ahmad Ridley, Senior Researcher, NSA
Ms. Sharothi Pikar, Associate Director for Cyber Strategies, OUSD(R&E), DoD