Gartner: 97% of companies did not achieve promised savings from AI

Executives promised that AI would sharply increase productivity and reduce costs; these same arguments were used to justify layoffs. Data from Gartner, presented at the HR Symposium in Orlando, shows that savings have hardly materialized. Only 3% of organizations exceed their cost reduction targets through productivity, while the remaining 97% do not achieve savings if they rely solely on employee efficiency. Meanwhile, 77% of board members are confident that AI will bring benefits in this way.

C.V. Viverito, Senior Director Analyst for HR at Gartner, believes that savings from AI will not occur now: value should be sought in applying the technology to what was previously impossible, such as reallocating employee time to new competencies and sources of growth. Nearly 4 out of 10 companies have already exceeded their AI budget, but 85% plan to increase spending in 2027.

The data aligns with other studies. In January 2026, PwC surveyed 4,454 executives from 95 countries: 56% saw neither cost reduction nor revenue growth. The MIT 2025 report showed that 95% of AI implementations do not yield profit growth. A 2024 survey revealed that generative AI can reduce productivity.

The key issue is the lack of a unified understanding of goals: in a survey of 297 HR directors, 78% admitted that leadership has not agreed on where AI creates value. Money is spent before a strategy is developed. Gartner recommends focusing investments where the impact is tangible, restructuring processes to fit the tools, and finding applications for the saved time: only 21% of freed hours were directed by employees towards more innovative work. This was also previously mentioned by Meta's CTO Andrew Bosworth.