According to the new 'ITBench-AA' benchmark by Artificial Analysis and IBM, even frontier AI models score below 50% on agentic enterprise IT tasks.

This indicates that AI agents are not yet ready for complex corporate processes.

AI-generated: summaries written by AI from the linked sources. How we use AI