{"id":"tp-25587","type":"article-newspaper","title":"Anthropic Says GLM-5.3 Can Build End-to-End Network Exploits","author":"Simon Yoon","publisher":"TOKENPOST Inc.","url":"https://www.tokenpost.com/news/technology/25587","datePublished":"2026-09-30T01:48:24.000Z","dateModified":"2026-09-30T01:48:24.000Z","language":"en","abstract":"Simple simulated attacks bypassed the model’s safeguards about 64% to 100% of the time, while protected Claude models resisted the same methods.","section":"Technology","references":[]}