Effectively evaluating AI models is more crucial than ever. But how do AI evaluations actually work? Our explainer lays out the different fundamental types of AI safety evaluations alongside their respective strengths and limitations.

Dewey Murdick and Miriam Vogel shared their expert analysis in an op-ed published by Fortune. In their piece, they highlight the urgent need for the United States to strengthen its AI literacy and incident reporting systems to maintain global leadership amid rapidly advancing international competition, especially from China’s booming AI sector.

A CSET report was highlighted in an article published by MIT Technology Review. The article discusses the U.S. military’s growing use of generative AI—such as chatbot-style tools modeled after ChatGPT—for intelligence analysis and decision support during deployments.

Artificial intelligence is reshaping military decision-making. This concise overview explores how AI-enabled systems can enhance situational awareness and accelerate critical operational decisions—even in high-pressure, dynamic environments. Yet, it also highlights the essential need for clear operational scopes, robust training, and vigilant risk mitigation to counter the inherent challenges of using AI, such as data biases and automation pitfalls. This report offers a balanced framework to help military leaders integrate AI responsibly and effectively.

Dewey Murdick and William Hannas shared their expert analysis in an op-ed published by The National Interest. In their piece, they discussed China's approach to artificial intelligence and the lessons it offers to American policymakers.

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.

Explainability and interpretability are often cited as key characteristics of trustworthy AI systems, but it is unclear how they are evaluated in practice. This report examines how researchers evaluate their explainability and interpretability claims in the context of AI-enabled recommendation systems and offers considerations for policymakers seeking to support AI evaluations.

In their op-ed in the Bulletin of the Atomic Scientists, Mia Hoffmann, Mina Narayanan, and Owen J. Daniels discuss the upcoming French Artificial Intelligence Action Summit in Paris, which aims to establish a shared and effective governance framework for AI.

This follow-up report builds on the foundational framework presented in the March 2024 CSET issue brief, “An Argument for Hybrid AI Incident Reporting,” by identifying key components of AI incidents that should be documented within a mandatory reporting regime. Designed to complement and operationalize our original framework, this report promotes the implementation of such a regime. By providing guidance on these critical elements, the report fosters consistent and comprehensive incident reporting, advancing efforts to document and address AI-related harms.