The resizing of images, which is typically a required part of preprocessing for computer vision systems, is vulnerable to attack. Images can be created such that the image is completely different at machine-vision scales than at other scales and the default settings for some common computer vision and machine learning systems are vulnerable.

Collaborating with allies to shape the trajectory of artificial intelligence and protect against digital authoritarianism

CSET Founding Director Jason Matheny testified before the House Budget Committee for its hearing, "Machines, Artificial Intelligence, & the Workforce: Recovering and Readying Our Economy for the Future." Dr. Matheny's full testimony as prepared for delivery can be found below.

Patent data can provide insights into the most active countries, fields and organizations in artificial intelligence research. This data brief analyzes worldwide trends in AI patenting to offer metrics on inventive activity.

CSET and the Bipartisan Policy Center partnered with Representatives Robin Kelly and Will Hurd to propose guidelines for national security considerations that must be addressed in a national AI strategy. The findings identify key areas for improvement in defense and intelligence to put the nation on a path to large-scale development and deployment of AI tools in promoting national security.

CSET submitted the following comment in response to the National Institute for Standards and Technology Plan for Federal Engagement in AI Standards Development.

Jeff Ding testified before the U.S.-China Economic and Security Review Commission for its hearing on "Technology, Trade, and Military-Civil Fusion: China’s Pursuit of Artificial Intelligence, New Materials, and New Energy."