Year: 2025
Type: preprint
Abstract: AI agents are AI systems that can achieve complex goals autonomously. Assessing the level of agent autonomy is crucial for understanding both their potential benefits and risks. Current assessments of autonomy often focus on specific risks and rely on run-time evaluations -- observations of agent actions during operation. We introduce a code-based assessment of autonomy that eliminates the need to run an AI agent to perform specific tasks, thereby reducing the costs and risks associated with run-time evaluations. Using this code-based framework, the orchestration code used to run an AI agent can be scored according to a taxonomy that assesses attributes of autonomy: impact and oversight. We demonstrate this approach with the AutoGen framework and select applications.
Source: arXiv (Cornell University)
Authors: Cihon, Peter, Stein, Merlin, Bansal, Gagan, Manning, Sam, Xu, Kevin
Language: English
Subfield: Safety Research
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W6979354677
Cites:
Cited by:
Related to: