ITBench-AA:在首个面向智能体企业IT任务的基准测试中,前沿模型得分低于50% ——由Artificial Analysis和IBM联合推出 AI资讯 · 7 阅读 · 2026-05-29 · 自动采集 ITBench-AA: Frontier Models Score Below 50% on the First Benchmark for Agentic Enterprise IT Tasks — by Artificial Analysis and IBM