Zhitao He
Titles as the sources gave them. In 1 entry, with 0 checked quotations.
In the Base
Every entry they appear in.
Multimodal-SEA: Benchmarking AI Agent Security under Multimodal Social Engineering Attacks
According to the abstract, the Multimodal-SEA authors introduce a benchmark of 608 tasks grounded in real fraud cases, and report that seven widely used multimodal models are all vulnerable to social engineering by AI agents, with attack success rates of 7.1% to 34.5%.