August 5, 2024
Some experts are concerned that AI chatbots could pose biological risks. But AI-enabled “biological tools” may pose greater risks and will likely require a distinct policy response.
January 22, 2025
Success at next month’s global AI summit in Paris hinges on three key outcomes: 1) ensuring the AI Summit Series' future, 2) securing senior US and Chinese engagement, and 3) demonstrating...
February 14, 2024
Recent AI progress has largely been driven by increases in the amount of computing power used to train new models. Governing compute could be an effective way to achieve AI policy goals...
February 17, 2025
Improvements in the capabilities of frontier AI systems present policymakers with a classic challenge: how to preserve the benefits of innovation while mitigating attendant risks. Certain...
Despite increasing interest in international AI benefit sharing, fundamental questions regarding motivations, mechanisms, and feasibility remain unexplored. In a new GovAI report, we describe three...
November 5, 2024
To succeed in a crowded international governance landscape, the AI Summit Series needs a clearly defined niche. Its defining traits should be that it: (a) invites companies and civil society to...
June 26, 2024
If policymakers could foresee how AI will impact labour markets, they could make better policy decisions. The nascent science of “automation evaluations” aims to provide this foresight.
May 15, 2024
AI agents could pose novel risks. Information about their activity in the real world could help us manage these risks.
April 9, 2024
A critical AI safety goal is understanding how new AI systems will behave in the real world. We can assess our understanding by trying to predict the results of model evaluations before running them.
March 4, 2024
The second AI Safety Summit is an opportunity to reinforce the world’s commitment to an ambitious summit series.
February 21, 2024
Falling development costs allow more and more groups to reproduce existing AI capabilities. But falling costs also benefit large compute investors, helping them maintain their leads by pushing...
December 17, 2023
This post aims to give policymakers an overview of available methods for preventing the misuse of general-purpose AI systems, as well as the limitations of these methods.
October 20, 2023
Developed countries shouldn't underestimate the value of including the Global South in international AI governance discussions.
August 30, 2023
A new report proposes creating an international body that can certify compliance with international standards on civilian AI.
August 8, 2023
Risks from the misuse of AI will continue to grow. While there are multiple ways to reduce misuse risks, restricting access to some AI capabilities will likely become increasingly necessary.
August 4, 2023
Voluntary commitments secured by the White House from leading AI labs are significant, but they also raise a number of questions.
July 10, 2023
The next generation of foundation models could threaten public safety. We explore the challenges of regulating frontier AI models and the building blocks of a potential regulatory regime.
July 9, 2023
The summit could produce a range of valuable outcomes. It may also be a critical and fleeting opportunity to bring China into global AI governance.
June 16, 2023
We propose an information-sharing regime between leading AI labs and the UK Office for Artificial Intelligence. This regime could support anticipatory policymaking and reduce risks.
June 8, 2023
Chinese policymakers are becoming interested in general AI, but see compute as a major challenge.
June 5, 2023
Our survey of 51 leading experts from AGI labs, academia, and civil society found overwhelming support for many AGI safety and governance practices.
April 17, 2023
Preliminary findings from a survey with over 13,000 people in 11 countries show that an overwhelming majority in Europe and the US there is a need for careful management of AI.
February 7, 2023
Numerous parties are calling for “the democratisation of AI”. But the phrase seems to refer to a variety of goals—which sometimes conflict.
April 11, 2022
The US National AI Research Resource should provide structured access to models, not just data and compute
January 20, 2022
There is a trend within AI research towards building large models that have a broad range of capabilities. Labs building such models face a dilemma when deciding how to share them.
April 13, 2021
Problems of cooperation—in which agents have opportunities to improve their joint welfare but are not easily able to do so—are ubiquitous and important.
