ITBench-AA：前沿模型在首个企业IT智能体任务基准测试中得分低于50%——由Artificial Analysis与IBM推出 AI资讯 · 5 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