How can we measure the reliability of machine learning systems? And do these measures really help us predict real world performance? A recent study by the Stanford Intelligent Systems Laboratory, supported by CSET funding, provides new evidence that models may perform well on certain reliability metrics while still being unreliable in other ways. This blog post summarizes the study’s results, which suggest that policymakers and regulators should not think of “reliability” or “robustness” as a single, easy-to-measure property of an AI system. Instead, AI reliability requirements will need to consider which facets of reliability matter most for any given use case, and how those facets can be evaluated.

On July 21, the White House announced voluntary commitments from seven AI firms to ensure safe, secure, and transparent AI. CSET’s research provides important context to this discussion.

Recent discussions of AI have focused on safety, reliability, and other risks. Lost in this debate is the real need to secure AI against malicious actors. This blog post applies lessons from traditional cybersecurity to emerging AI-model risks.

Machine learning advances have powered the development of new and more powerful generative language models. These systems are increasingly able to write text at near human levels. In a new report, authors at CSET, OpenAI, and the Stanford Internet Observatory explore how language models could be misused for influence operations in the future, and they provide a framework for assessing potential mitigation strategies.