Kao introduces Kao-MIKOMIL AI agent to support demand planning
In an October 13, 2026 announcement, Kao Corporation introduced Kao-MIKOMIL, an AI agent that turns forecasts from its in-house demand forecasting model into demand plan proposals by incorporating planned promotions, sales channels, pricing and other business measures. Introduced for the KATE cosmetics brand in July 2026, it is intended to expand gradually to other cosmetics brands. The agent explains the rationale for its proposals; business division staff review them and make the final decision. Kao links data across business, production, logistics and sales and targets approximately 25% lower cosmetics product inventory by 2027 versus 2024. This is a company-wide objective, not a measured result attributable to the agent.
Illustrative warehouse photograph, not a Kao facility or Kao-MIKOMIL interface. Photo: Tiger Lily / Pexels. Source: https://www.pexels.com/photo/man-holding-digital-tablet-4484155/ . License: https://www.pexels.com/license/
Kao-MIKOMIL moves AI beyond generating a statistical forecast into proposing a demand plan that incorporates commercial assumptions. The consequential shift is at the decision interface: a proposal and its rationale are made available to business staff, who retain accountability for the final plan.
For demand-planning leaders, value depends on visible assumptions, meaningful planner review, controlled overrides and clarity about who authorizes the final demand plan. The failure mode is a plausible AI recommendation being accepted without sufficient business challenge, especially when promotions, channels or prices are uncertain. Kao's inventory reduction goal is an ambition of its broader planning programme, not evidence of value already delivered by this agent.