FraudBench announced to evaluate AI deception
A new dataset based on 159 scenarios has been introduced to evaluate the propensity of LLMs to recommend, assist, or directly act in ways that constitute fraud under U.S. law.
FraudBench has been introduced to evaluate the propensity of large language models (LLMs) to recommend, assist, or directly take actions that would constitute fraud under U.S. law. This was announced on Hugging Face Papers.
The evaluation contains a novel dataset of 159 scenarios based on real-world U.S. court cases and validated by legal experts.
Each scenario is presented to LLMs via three tasks at varying levels of autonomy: advising users, drafting messages for users to send, and acting directly on behalf of users, totaling 477 tasks. It received 1 upvote on Hugging Face Daily Papers.
This item is a summary. The full, checkable version is at the original source:
Read the original at Hugging Face Papers ↗