This year, CSET researchers returned to the DEF CON cybersecurity conference to explore how understandings of AI red-teaming practices have evolved among cybersecurity practitioners and AI experts. This blog post, a companion to "How I Won DEF CON’s Generative AI Red-Teaming Challenge", summarizes our takeaways and concludes with a list of outstanding research questions regarding AI red-teaming, some of which CSET hopes to address in future work.

In August 2024, CSET Research Fellow Colin Shea-Blymyer attended DEF CON, the world’s largest hacking convention to break powerful artificial intelligence systems. He participated in the AI red-teaming challenge, and won. This blog post details his experiences with the challenge, what it took to win the grand prize, and what he learned about the state of AI testing.

Since 2022, U.S. export controls have restricted the highest-performing AI chips from being exported to China. The Biden administration likely did not intend to control CPUs (i.e., general-purpose processors) with these restrictions. However, CPUs are increasingly subject to export controls because chip designers are incorporating specialized elements for AI computation into CPUs. In this blog post, we discuss the implications of controlling AI-capable CPUs and make recommendations for the Bureau of Industry and Security (BIS) at the U.S. Department of Commerce.

The National Artificial Intelligence Research Resource (NAIRR) pilot provides federal infrastructure, including computational resources, to U.S. AI researchers. This blog post estimates the compute provided through the pilot’s initial six resources. We find that the total compute capacity of the initial resources is roughly 3.77 exaFLOPS, the equivalent of approximately 5,000 H100 GPUs (using the tensor cores optimal for AI). Factoring in the amount of time these resources are available for use, we find that the overall compute allocated is roughly 3.26 yottaFLOPs. The pilot is a significant first step in providing compute to under-resourced organizations, although it is a fraction of what is available to industry.

Over the past year, artificial intelligence has quickly become a focal point in K-12 education. This blog post describes new and existing K-12 AI education efforts so that U.S. policymakers and other decision-makers may better understand what’s happening in practice.

The Carnegie Classification of Institutions of Higher Education is making changes to drastically simplify the criteria that determine its highly coveted R1 top-tier research classification. Last year, CSET Senior Fellow, Jaret Riddick, wrote about a new law from Congress, Section 223 of the 2023 National Defense Authorization Act, intended to leverage existing Carnegie classification criteria to increase defense research capacity for historically Black colleges and universities. Now, research is needed to understand how the changes proposed for 2025 classification criteria impact U.S. Department of Defense goals for eligible HBCU partners.

This blog post assesses how different priorities can change the risk-benefit calculus of open foundation models, and provides divergent answers to the question of “given current AI capabilities, what might happen if the U.S. government left the open AI ecosystem unregulated?” By answering this question from different perspectives, this blog post highlights the dangers of hastily subscribing to any particular course of action without weighing the potentially beneficial, risky, and ambiguous implications of open models.

As the U.S. government considers banning genomics companies from China in the Biosecure Act, it opens a broader question of how the U.S. and other market economies should deal with China’s national champions. This blog post provides an overview of BGI and how China’s industrial policy impacts technology development.

As the U.S. government tightens its controls on China’s semiconductor ecosystem, a new dimension is increasingly worrying Congress: the open-source chip architecture known as RISC-V (pronounced “risk-five”). This blog post provides an introduction to the RISC-V architecture and an explanation of what policy-makers can do to address concerns about this open architecture.

This guide provides a run-down of CSET’s research since 2019 for first-time visitors and long-term fans alike. Quickly get up to speed on our “must-read” research and learn about how we organize our work.