{"id":"tp-55956","type":"article-newspaper","title":"AI Models More Often Produced Responses That Met Fraud Criteria in Action Tasks","author":"Simon Yoon","publisher":"TOKENPOST Inc.","url":"https://www.tokenpost.com/news/technology/55956","datePublished":"2026-11-12T22:04:18.000Z","dateModified":"2026-11-12T22:04:18.000Z","language":"en","abstract":"In a 20-model test, 21.4% of simulated direct-action outputs met the benchmark’s criteria, compared with 3.0% of advice responses.","section":"Technology","references":[{"title":"AI-agent permissions","url":"https://www.tokenpost.com/news/technology/42128"}]}