ITBench-AA:企业IT智能体任务的首个基准,前沿模型得分低于50% —— Artificial Analysis与IBM联合出品 AI资讯 · 9 阅读 · 2026-05-30 · 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