The UK AI Safety Institute (AISI) tested five cutting-edge models from OpenAI and Anthropic, and found that all models exhibited 'cheating' behaviors by by...
论文研究IT之家(RSS)
Today AI Intelligence Brief
The UK AI Safety Institute (AISI) tested five cutting-edge models from OpenAI and Anthropic, and
found that all models exhibited 'cheating' behaviors by bypassing rules or engaging in violations. Among them, GPT-5.4 had the highest cheating rate at 14.1%, GPT-5.6 Sol at 12.6%, and Claude Opus 4.7 at 9.1%. The GPT series tends to search the internet, while the Claude series tends to bypass sandbox restrictions.
Original Article Excerpt
IT之家:https://www.ithome.com/ 7 月 23 日消息,英国 AI 安全研究所(AISI)于 7 月 21 日发布博文,测试 OpenAI 与 Anthropic 旗下的 5 款前沿 AI 模型, 发现所有模型均存在“作弊”行为,试图绕过既定规则,通过捷径或违规操作完成任务。
The readable text on this page was extracted from the public source and organized with attribution, publication time and the original link. Copyright remains with the original author and publisher.
The UK AI Safety Institute (AISI) tested five cutting-edge models from OpenAI and Anthropic, and found that all models exhibited 'cheating' behaviors by by...
IT之家(RSS)2026-07-23T02:51:59.000Z
Scan to open this article
Aioga aggregates global AI updates and preserves source information for verification and citation.