Story

On air. The free wire runs 15 minutes behind live. See pricing

World, crypto, finance and tech news, scored for impact.

AITokenPostKorean, translatedFiled 38 h ago

An estimated 6–11% of AI agent tasks were high-complexity.

Original headlineAI 에이전트 업무 중 6~11%가 고복잡도로 추정됐다

Open the sourceReported by TokenPost, tokenpost.kr, 16 November 2026, by 최윤서

Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.

A Wharton School and Accenture report estimates that only 6–11% of the tasks classified as suitable for AI agents, by industry, are high-complexity work requiring frontier models. The study, "What Does Agentic AI Really Cost?", reviewed 923 occupations and more than 18,000 tasks and activities from O*NET, narrowing to 11,339 tasks deemed agent-eligible after accounting for human interaction and physical activity needs. It recommends that enterprises assess task difficulty alongside data and system requirements to choose models at the design stage rather than defaulting to the most capable ones. The findings matter because they suggest most agentic automation can run on cheaper, smaller models, with cost tied to workflow, regulatory, integration and sensitive-data constraints. What to watch next is how firms segment workloads by complexity and whether model selection at design time becomes standard practice as agent deployments scale.

Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.

Named in this story

Wharton School, Accenture (ACN)

A linked name opens that asset's sentiment page.