UK and Irish businesses are facing higher generative AI costs as deployments move from pilots into regular operations, according to research commissioned by SAS. A survey of 100 senior enterprise technology decision-makers found 22% of organisations have fully integrated generative AI into regular processes, up from 9% in 2024.
Cost concerns are most acute among mature users. Within the fully integrated group, 41% said prohibitive large language model costs were an active barrier to implementation, compared with 32% across the wider market. Returns remain uneven, with 45% of fully integrated organisations reporting below-expected return on investment, even as nearly half of enterprises actively using generative AI reported significant operating cost and time savings.
SAS said consumption-based pricing is complicating forecasts because usage rises as more workflows, user groups, and automated tasks are added. Agentic AI could intensify that pressure, with 30% of surveyed organisations already investigating or piloting systems for tasks such as case handling, onboarding, and investigations. SAS cited Gartner analysis that agentic AI models can require five to 30 times more tokens per task than standard generative AI, with a single workflow making 10 to 20 separate AI calls to complete one task. SAS urged businesses to introduce governance, budgets, usage controls, and full-deployment cost stress tests before scaling.
