MS lowered sales targets for AI products including agents by department after sales of such products failed to meet targets in fiscal year 2025, which ended last June, reported American information technology (IT) specialized media The Information, citing internal sources on Dec. 3 (local time).

The particular problem was the “foundry” product, which helps corporate customers develop their own AI apps and agents.

One cloud sales division set a goal to increase sales of this product by 50% and encouraged sales staff to promote sales, but after tabulating results following the fiscal year-end, the rate of achieving quotas was less than one-fifth.

Other business divisions also set goals to double sales targets for the same product but failed to achieve them.

Ultimately, these business divisions lowered their sales targets for the current fiscal year that began last July to 25-50% compared to the previous year.

Industry analysts explained that such measures by MS to lower targets for specific products are unusual.

Such poor sales of AI agents stem from corporate customers hesitating to adopt these products.

While agents are said to be capable of performing tasks on behalf of humans, it is difficult to accurately measure the cost-saving effects that result from their utilization.

Additionally, some companies have safety concerns as even minor mistakes or malfunctions in areas such as cybersecurity work or financial automation can lead to significant losses.

Private equity firm Carlyle experienced difficulties after introducing MS’s AI Copilot last year for meeting summaries and financial model creation.

The AI failed to properly extract data from external apps. Eventually, Carlyle recently reduced spending on Copilot tools.

Regarding this, an MS spokesperson told U.S. business broadcaster CNBC, “The total sales allocation for AI products has not been adjusted downward.”

Investment bank D.A. Davidson analyst Gil Luria explained to Reuters regarding the report, “The industry is currently in the early stages of AI adoption,” adding “This doesn’t mean AI products don’t help improve corporate productivity, but rather that it may be more difficult than they thought.”

Source: BusinessKorea