ITBench-AA:前沿模型在首个企业自主IT任务基准测试中得分低于50%——由Artificial Analysis和IBM联合发布 AI资讯 · 5 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