Sort by
Refine Your Search
-
research initiative funded by ARIA, titled Aggregating Safety Preferences for AI Systems: A Social Choice Approach. The project operates at the interface of AI safety and computational social choice, and
-
postdoctoral research associate (PDRA) position on the project project Aggregating Safety Preferences for AI Systems: A Social Choice Approach, funded by ARIA under the Safeguarded AI TA1.4 call. The project
-
The Department of Earth Sciences, University of Oxford, seeks to appoint a full-time Postdoctoral Research Assistant in Security of Carbon Dioxide Drawdown During Enhanced Rock Weathering. The post
-
to the 4th February 2026. You will be investigating the safety and security implications of large language model (LLM) agents, particularly those capable of interacting with operating systems and external APIs
-
. Due to the nature of the research at the Kennedy Institute of Rheumatology, this job will require additional security pre-employment checks: • A satisfactory basic Disclosure and Barring Service
-
management of sport injuries, with emphasis on safety rather than performance. You will be part of an interdisciplinary team of pioneering researchers, with the primary aim to develop cutting-edge robotics and
-
translate research outcomes to real world impact. The post holder will be required to carry out research in the area of cell characterisation and safety by examining the thermal and electrical responses
-
security pre-employment checks: • A satisfactory enhanced Disclosure and Barring Service check • University security screening (eg identity checks) The closing date for applications is 12 noon
-
DPhil students, manage data analysis pipelines, and contribute to publications and grant writing. This post is ideally suited to someone aiming to secure a long-term fellowship and build an independent
-
fixed-term to the 28th Feb 2027. There has been massive recent interest and efforts into systematic evaluation of the safety of LLM & VLM agents, but these works have focused exclusively on single-agent