ITBench-AA:前沿模型在首个面向智能体的企业IT任务基准中得分低于50% —— 由Artificial Analysis与IBM联合推出 AI资讯 · 7 阅读 · 2026-06-01 · 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