Large US companies are struggling to move artificial intelligence from experimental pilots to autonomous action, according to a new report from PYMNTS Intelligence. The findings suggest that while organizations have cleared the initial hurdle of approving AI projects, they face significant friction in granting agents the authority to execute tasks independently.
The report, titled “The Agentic Edge Index: How Leading Companies Put AI to Work,” introduces a new benchmark designed to measure how deeply AI supports daily operations and the level of autonomy granted to these systems. The data indicates that widespread independent execution remains rare, even among the most advanced organizations.
Low Scores Across the Board
The Agentic Edge Index evaluates AI usage across 75 distinct activities spanning eight business areas. The scoring model weighs activities with greater business impact more heavily, aiming to provide executives with a clear view of their progress. The results show a significant gap between ambition and reality.
Nine in 10 enterprises scored less than 25 on the index’s 100-point scale. This low average suggests that most large corporations are still in the early stages of integrating AI into their workflows. Even the companies furthest ahead have not yet achieved widespread independent execution, indicating that the transition from assistance to action is far from complete.
The research draws on 28 surveys conducted with 1,687 executives at US enterprises with at least $1 billion in annual revenue. The Index comparisons specifically utilize responses from 366 AI decision-makers surveyed in 2026, providing a current snapshot of the corporate landscape.
People and Process Outweigh Technology
Reaching higher levels of AI maturity requires more than purchasing new software. Just over half of executives surveyed stated that people, processes, and company culture constrain AI adoption more than the technology itself. This insight shifts the focus from technical infrastructure to organizational readiness.
Executives are increasingly recognizing that training staff and establishing clear responsibilities are critical components of an investment plan. Without these foundational elements, AI agents cannot operate effectively within existing business structures. The report suggests that companies must address these human and procedural gaps to unlock the full potential of agentic AI.
The Agentic Edge Index serves as a tool for leaders to assess their own progress against industry peers. By highlighting where AI support is deepest and where authority is limited, the benchmark helps identify the next areas for development. The data underscores that achieving the “agentic edge” is a complex challenge involving cultural shift as much as technological upgrade.
Source: PYMNTS
