Opening article
The AISI Blog has published a safety case template focused on inability arguments. The resource provides guidance on how to write a section of a safety case demonstrating that an AI system lacks offensive cyber capabilities. The template is designed to help structure these specific technical safety arguments.
Note to readers: we changed our name to the AI Security Institute on 14 February 2025. Read more here.
Safety cases are clear, assessable arguments that show that a system is safe in a given context. At the AI Safety Institute, we think safety cases could provide an approach for AI safety that works for both current and much higher-capability systems. Our new paper , written in collaboration with the Centre for the Governance of AI, provides a safety case template that could be used for some current AI systems.
Safety cases are used across the automotive (including autonomous vehicles), civil aviation, defence, nuclear, petrochemical, rail and healthcare industries ( Favaro et al., 2023 ; Sujan et al., 2016 ; Bloomfield et al., 2012 ; Inge, 2007 ). In the case of frontier AI, two developers have highlighted safety cases as part of their approaches to safety: Google DeepMind propose to use safety cases for future systems as part of their frontier safety framework , while Anthropic noted that the recent update to their responsible scaling policy was inspired by safety case methodologies.
For safety cases to be successful in frontier AI, they need to be both practical and reliable. We don’t yet know the best way to write safety cases for frontier AI systems. We’ll need to build on best practice from other industries, while also tailoring techniques to AI’s unique circumstances.
But if we can find good ways of making these arguments, safety cases may become a robust but flexible tool assuring the safety of frontier AI systems.
Our new template shows that it’s possible to write safety cases for current systems, by bringing together existing techniques.
This resource provides practical guidance for AI developers and safety researchers working to demonstrate that their models do not possess dangerous offensive cyber capabilities. Such templates help standardize safety evaluations across the industry.