Unlike other domains of conflict, and unlike other fields with high anticipated risk from AI, the cyber domain is intrinsically digital with a tight feedback loop between AI training and cyber application. Cyber may have some of the largest and earliest impacts from AI, so it is important to understand how the cyber domain may change as AI continues to advance. Our approach reviewed the literature, collecting nine arguments that have been proposed for offensive advantage in cyber conflict and nine proposed arguments for defensive advantage.

As new advanced AI systems roll out, there is widespread disagreement about malicious use risks. Are bad actors likely to misuse these tools for harm? This report presents a simple framework to guide the questions researchers ask—and the tools they use—to evaluate the likelihood of malicious use.

In response to the Office of Science and Technology Policy's request for input on an AI Action Plan, CSET provides key recommendations for advancing AI research, ensuring U.S. competitiveness, and maximizing benefits while mitigating risks. Our response highlights policies to strengthen the AI workforce, secure technology from illicit transfers, and foster an open and competitive AI ecosystem.

Large generative models are widely viewed as the most promising path to general (human-level) artificial intelligence and attract investment in the billions of dollars. The present enthusiasm notwithstanding, a chorus of ranking Chinese scientists regard this singular approach to AGI as ill-advised. This report documents these critiques in China’s research, public statements, and government planning, while pointing to additional, pragmatic reasons for China’s pursuit of a diversified research portfolio.

Maintaining U.S. technological leadership in the years ahead will require policymakers to promote competition in the AI market and prevent industry leaders from wielding their power in harmful ways. This brief examines trends in U.S. mergers and acquisitions of artificial intelligence companies. The authors found that AI-related M&A deals have grown significantly over the last decade, with large U.S. tech companies being the most prolific acquirers of AI firms.

Artificial intelligence models have become increasingly adept at generating computer code. They are powerful and promising tools for software development across many industries, but they can also pose direct and indirect cybersecurity risks. This report identifies three broad categories of risk associated with AI code generation models and discusses their policy and cybersecurity implications.

The Chinese Academy of Sciences is among the most important S&T organizations in the world and plays a key role in advancing Beijing’s S&T objectives. This report provides an in-depth look into the organization and its various functions within China’s S&T ecosystem, including advancing S&T research, fostering the commercialization of critical and emerging technologies, and contributing to S&T policymaking.

Computer scientists have long sought to build systems that can actively and autonomously carry out complicated goals in the real world—commonly referred to as artificial intelligence "agents." Recently, significant progress in large language models has fueled new optimism about the prospect of building sophisticated AI agents. This CSET-led workshop report synthesizes findings from a May 2024 workshop on this topic, including what constitutes an AI agent, how the technology is improving, what risks agents exacerbate, and intervention points that could help.

As critical infrastructure operators and providers seek to harness the benefits of new artificial intelligence capabilities, they must also manage associated risks from both AI-enabled cyber threats and potential vulnerabilities in deployed AI systems. In June 2024, CSET led a workshop to assess these issues. This report synthesizes our findings, drawing on lessons from cybersecurity and insights from critical infrastructure sectors to identify challenges and potential risk mitigations associated with AI adoption.

A core question in policy debates around artificial intelligence is whether federal agencies can use their existing authorities to govern AI or if the government needs new legal powers to manage the technology. The authors argue that relying on existing authorities is the most effective approach to promoting the safe development and deployment of AI systems, at least in the near term. This report outlines a process for identifying existing legal authorities that could apply to AI and highlights areas where additional legislative or regulatory action may be needed.