robot hand holding earth (vectorfusionart/

Cyber defenses for machine learning

The National Science Foundation's  Secure and Trustworthy Cyberspace (SaTC) program has recently invested  $78.2 million in 225 new projects in 32 states for  research and education on artificial intelligence, cryptography, network security, privacy and usability.

The largest of the projects is $9.98 million grant to establish the Center for Trustworthy Machine Learning (CTML) at Penn State College of Engineering. The multi-institution, multi-disciplinary center will focus on understanding the security risks of machine learning and devise the tools, metrics and methods to manage and mitigate security vulnerabilities.

The new CTML will work to develop an arsenal of defensive techniques to build future systems in a safer, more secure manner. 

"Machine learning is fundamentally changing the way we live and work -- from autonomous vehicles, digital assistants, to robotic manufacturing -- we see computers doing complex reasoning in ways that would be considered science fiction just a decade ago," said Patrick McDaniel, lead principal investigator of the CTML project. "We have a unique opportunity at this time, before machine learning is widely deployed in critical systems, to develop the theory and practice needed for robust learning algorithms that provide rigorous and meaningful guarantees."

CTML is a collaboration with Stanford University, University of California-Berkeley, University of California-San Diego and the University of Wisconsin-Madison.


Charter Sponsors