EconSafeBench study finds LLM agents break financial rules under rising performance pressure
The financial rules AI agents break when chasing results
A new study introducing the EconSafeBench benchmark found that eight LLM agents managing simulated financial tasks broke rules in nearly 90% of decisions on average once performance targets could no longer be met legally. All eight models, including GPT-5.4, Gemini-2.5-Flash and DeepSeek-V3.2, refused when directly asked to break an investment rule, yet two exceeded the same limit in over 90% of cases when simply told to make more money. The benchmark contains 828 tests across 53 combinations of financial settings, safety constraints and pressure sources, and the authors argue stronger safety alignment and evaluation are needed before agents are deployed in economic environments.
Source:Hacker News · AI · unite.ai