New research from The Harris Poll and Hyperscience reveals how organizations are rethinking GenAI as cost, data quality, and operational complexity reshape the path to scale.
The first wave of enterprise GenAI strategy was largely about capability: find the biggest, most powerful model and put it to work.
Now the reality of running GenAI at scale is setting in.
86% of organizations say managing GenAI compute and token costs is now as critical to technology strategy as managing cloud costs. 52% say their GenAI infrastructure costs have already come in higher than expected. And 80% are moving away from relying on a single large model.
The new question isn’t whether organizations will invest in GenAI. It’s how they can scale it without sacrificing cost, accuracy, speed, or control.
Download the new research from The Harris Poll and Hyperscience to explore how organizations are responding to the economics and operational realities of GenAI and what the shift means for the next generation of AI infrastructure.
Inside the report, you’ll discover:
- Why the one-model-fits-all approach is fading: Only 19% of organizations expect to rely on a single large foundational model over the next 12–18 months, while 62% are optimizing model selection and routing.
- How token economics are reshaping AI strategy: 64% have implemented hard cost caps on AI infrastructure spending, and 47% are reevaluating their GenAI strategy because of cost.
- Why data, not models, is becoming the bottleneck: 86% say data quality issues are hurting the performance of their GenAI applications, while 56% describe their data extraction and preparation processes as inefficient or manual.
- Where organizations are turning for help: 77% rely on outside vendors for AI routing and orchestration rather than building those capabilities entirely in-house.
- Why GenAI investment isn’t slowing: Despite cost pressure and operational challenges, 78% plan to increase their GenAI investment over the next 12 months.
The research at a glance
86%
Say GenAI cost management is as critical as cloud cost management
80%
Moving away from a single-large-model strategy
78%
Plan to increase GenAI investment
86%
Say data quality is hurting GenAI performance
Download the report to learn more.