New benchmark reveals leading AI models struggle with real IT troubleshooting

Original: ITBench-AA: Frontier Models Score Below 50% on the First Benchmark for Agentic Enterprise IT Tasks — by Artificial Analysis and IBM

Source: huggingface.co ↗