Artificial intelligence will play an increasingly important role in cyber defense, but vulnerabilities in AI systems call into question their reliability in the face of evolving offensive campaigns. Because securing AI systems can require trade-offs based on the types of threats, defenders are often caught in a constant balancing act. This report explores the challenges in AI security and their implications for deploying AI-enabled cyber defenses at scale.

Cybersecurity operators have increasingly relied on machine learning to address a rising number of threats. But will machine learning give them a decisive advantage or just help them keep pace with attackers? This report explores the history of machine learning in cybersecurity and the potential it has for transforming cyber defense in the near future.

Cyber AI Director Ben Buchanan sat down with Scientific American to discuss the cyberattack on an Oldsmar, Florida water supply facility and how to protect public facilities in the future.

Open source intelligence (OSINT) and science and technology intelligence (STI) are realized differently in the United States and China, China putting greater value on both. In the United States’ understanding, OSINT “enables” classified reporting, while in China it is the intelligence of first resort. This contrast extends to STI which has a lower priority in the U.S. system, whereas China and its top leaders personally lavish great attention on STI and rely on it for national decisions. Establishing a “National S&T Analysis Center” within the U.S. government could help to address these challenges.

As states turn to AI to gain an edge in cyber competition, it will change the cat-and-mouse game between cyber attackers and defenders. Embracing machine learning systems for cyber defense could drive more aggressive and destabilizing engagements between states. Wyatt Hoffman writes that cyber competition already has the ingredients needed for escalation to real-world violence, even if these ingredients have yet to come together in the right conditions.

Machine learning may provide cyber attackers with the means to execute more effective and more destructive attacks against industrial control systems. As new ML tools are developed, CSET discusses the ways in which attackers may deploy these tools and the most effective avenues for industrial system defenders to respond.

Chris Rohlf is a Non-Resident Research Fellow at CSET, where he works on the CyberAI Project.

Machine learning advances are transforming cyber strategy and operations. This necessitates studying national security issues at the intersection of AI and cybersecurity, including offensive and defensive cyber operations, the cybersecurity of AI systems, and the effect of new technologies on global stability. 

Accelerating threats to cybersecurity, the impact of automation on cyber defense, and the degree to which cyber operations will become faster and more powerful are among the subjects that CSET will now start to explore thanks to a grant from the William and Flora Hewlett Foundation.