Body
Major consulting firms and government agencies are mandating generative AI usage for staff and tying adoption metrics to promotion tracks, even as internal confusion over implementation and expected returns leaves employees and mid-level managers uncertain about the technology’s workplace role. Executives are rolling out dashboards to track usage rates and demanding higher adoption percentages, while workers and mid-level managers report that technology decisions are frequently divorced from practical operational needs.
In February, global consultancy Accenture reportedly told staff that promotions to top roles would require regular adoption of AI tooling, and the firm would track employee usage of its internally developed AI platform. In May, rival firm KPMG said it developed a dashboard to track whether its U.S. employees meet a 75% usage target for its AI tools. KPMG said the tracking is part of a holistic effort to help people move up the AI maturity curve.
The push toward mandated generative AI use contrasts with technical assessments from engineers on the ground. An AI engineer named Malcolm, whose real name was not used, said executives at a data analysis firm wanted to use generative AI to categorize a customer database into a range of personas. Malcolm advised against the approach, arguing that a traditional machine learning model would produce more consistent and repeatable results at a lower cost. Management proceeded with generative AI anyway. The result was a process that Malcolm said was less accurate and more expensive, but allowed the organization to say it was embracing the technology.
Government agencies face similar implementation headwinds. The U.K. government is planning to use AI to help rewire state operations and boost efficiency across Whitehall, but the FDA civil servant union found that the rollout strategy lacks worker input. The union reported that fewer than one-third of civil servants were consulted on how the technology could be deployed. FDA general secretary Dave Penman said the rollout is inconsistent across departments, which limits the potential productivity gains. Penman said the change is being done to workers rather than with them.
Consultants observe a recurring disconnect between executive rhetoric and operational purpose. Dan Boyles, CEO of Hello AI Collective, said organizations are quick to highlight AI adoption but unclear on why they are adopting it or how they expect to benefit. Boyles described sitting with the C-suite of an oil and gas company asking for the reason behind the push for AI. The CEO cited the need to keep up with competitors, the head of sales said the goal was to make more money, and the marketing team wanted to stop using outside contractors. Boyles said none of them could agree on a single objective.
Undefined strategic goals correlate with poor investment returns, according to a senior consultant at a large consulting firm who spoke on the condition of anonymity. The consultant said the wreckage of unfocused rollouts consists of organizations not receiving the expected return on investment and failing to secure employee engagement. The consultant’s firm provides everyone with access to two AI tools and allows staff to request specialist tools for specific tasks like coding. If a role demands it, some employees have access to four or five AI tools. Organizations must also consider the people side of the equation, the consultant said, noting there are generational differences in confidence levels regarding the technology, as well as potential gender differences.
Before staff gain access to any tool, they must complete mandatory training covering AI ethics and risks, including system bias. The training also makes clear that AI tools can be sycophantic and prone to hallucination, the senior consultant said.
Organizational culture heavily influences the success of these deployments. Caroline Rawlinson, CEO of Culture Amp, which tracks employee experiences, said AI tends to accelerate pre-existing cultural dynamics for better or worse. Culture Amp found that while nine out of 10 human resources professionals expect to increase their use of generative AI, one-third said no single executive currently owns AI strategy at their companies. Rawlinson said that placing AI technology on top of a fragmented or fear-based culture guarantees failure.
Rawlinson said the best-case scenario for poorly integrated rollouts is a slow deployment where employees do not understand what they are being asked to achieve or how to use the provided tools. The worst case, she said, is a large, wasted effort. The FDA union survey reflects these cultural friction points, showing that while civil servants were generally open to using AI to improve productivity, research indicated doubt that management is equipped to handle the transformation. In the oil and gas example, Boyles said the president eventually clarified the core motivation: he wanted to increase operating earnings because he plans to sell the company. That clarification allowed consultants to work through departmental processes, identify bottlenecks, and determine where AI could actually assist.