Publications

CSET produces evidence-driven analysis in a variety of forms, from informative graphics and translations to expert testimony and published reports. Our key areas of inquiry are the foundations of artificial intelligence — such as talent, data and computational power — as well as how AI can be used in cybersecurity and other national security settings. We also do research on the policy tools that can be used to shape AI’s development and use, and on biotechnology.

Analysis

AI Safety and Automation Bias

Lauren Kahn, Emelia Probasco, and Ronnie Kinoshita
| November 2024

Automation bias is a critical issue for artificial intelligence deployment. It can cause otherwise knowledgeable users to make crucial and even obvious errors. Organizational, technical, and educational leaders can mitigate these biases through training, design, and processes. This paper explores automation bias and ways to mitigate it through three case studies: Tesla’s autopilot incidents, aviation incidents at Boeing and Airbus, and Army and Navy air defense incidents.

Applications


Workforce


Filter publications

See our original translation of China's 14th Five-Year Plan for the Development of the Big Data Industry, covering the period through 2025.

CHIPS for America Act funding will result in the construction of new semiconductor fabrication facilities (“fabs”) in the United States, employing tens of thousands of workers. This policy brief assesses the occupations and backgrounds that will be most in-demand among new fabs, as well as options for ensuring availability of the necessary talent. Findings suggest the need for new immigration pathways for experienced foreign fab workers, and investments in workforce development.

Data Brief

Exploring Clusters of Research in Three Areas of AI Safety

Helen Toner and Ashwin Acharya
| February 2022

Problems of AI safety are the subject of increasing interest for engineers and policymakers alike. This brief uses the CSET Map of Science to investigate how research into three areas of AI safety — robustness, interpretability and reward learning — is progressing. It identifies eight research clusters that contain a significant amount of research relating to these three areas and describes trends and key papers for each of them.

Data Visualization

Classifying AI Systems

Catherine Aiken and Brian Dunn
| December 2021

​​This Classifying AI Systems Interactive presents several AI system classification frameworks developed to distill AI systems into concise, comparable and policy-relevant dimensions. It provides key takeaways and framework-specific results from CSET’s analysis of more than 1,800 system classifications done by survey respondents using the frameworks. You can explore the frameworks and example AI systems used in the survey, and even take the survey.

See our original translation of a an article summarizing China's 2022 Nationwide S&T Work Conference, which reviewed China's progress toward its technological goals during 2021 and set new priorities for 2022.

See our original translation of China's 14th Five-Year Plan for Promoting the High-Quality Development of the National Standards System, covering the period through 2025.

Since 1990, the U.S. share of global semiconductor manufacturing capacity has declined while the shares of China, South Korea, and Taiwan have increased. If carefully targeted, CHIPS for America Act incentives could reverse this trend for the types of chips that matter most to U.S. national security. In this policy brief, the author assesses how CHIPS Act incentives should be distributed across different types of chips.

Data Brief

Chinese and U.S. University Rankings

Jack Corrigan and Simon Rodriguez
| January 2022

The strength of a country’s talent pipeline depends in no small part on the quality of its universities. This data brief explores how Chinese and U.S. universities perform in two different global university rankings, why their standings have changed over time, and what those trends mean for graduates.

Analysis

AI and Compute

Andrew Lohn and Micah Musser
| January 2022

Between 2012 and 2018, the amount of computing power used by record-breaking artificial intelligence models doubled every 3.4 months. Even with money pouring into the AI field, this trendline is unsustainable. Because of cost, hardware availability and engineering difficulties, the next decade of AI can't rely exclusively on applying more and more computing power to drive further progress.

See our original translation of a 2021 plan from a local PRC government outlining its strategy for implementing military-civil fusion.