Companies Mismeasure AI Impact, Driving Cost Overruns

Major corporations are struggling to measure the business value of artificial intelligence, with many defaulting to flawed metrics that reward high activity over meaningful outcomes, according to recent reports and expert analysis. This measurement problem is leading to misaligned incentives, budget overruns, and a failure to capture the technology's full strategic potential.

A prominent example is Uber, where engineering teams exhausted their entire 2026 AI budget by April after the company ranked engineers on internal leaderboards based on tool usage. The company's chief technology officer acknowledged the need to reassess the approach, as the tools proved "too successful to afford at scale." Similarly, Starbucks has tied a quarter of some tech workers' bonuses to department-wide AI adoption goals, defined as using an AI assistant multiple times per week. Critics argue such metrics incentivize the appearance of adoption rather than genuine productivity gains.

The core issue, experts say, is that many organizations are measuring inputs-such as token consumption, time spent in AI tools, or frequency of use-instead of outputs tied to business results. An engineer who uses an AI...