AI Rollouts Sputter Amidst Confusion and Lack of Staff Consultation

Efforts by companies and governments to integrate artificial intelligence are facing significant hurdles, characterized by a lack of clear strategic objectives and inadequate consultation with staff, according to recent reports.

The UK government, for instance, is pursuing AI to enhance efficiency across Whitehall. However, research from the civil servant union, the FDA, indicates that while civil servants are open to AI for productivity gains, there is skepticism regarding management's ability to lead such a transformation. The union found that less than a third of civil servants were consulted on AI implementation, leading to a feeling that "change is being done to workers, not with them." FDA general secretary Dave Penman noted that the inconsistent rollout across different government departments is limiting potential productivity benefits.

Experts observe a pattern where organizations are quick to announce AI adoption without articulating the specific reasons or expected benefits. Dan Boyles, CEO of consultancy Hello AI Collective, recounted an instance with an oil and gas company where the C-suite could not agree on the rationale for using AI. While the CEO cited competitive pressure, the head of sales aimed for increased revenue, and the marketing team sought to reduce reliance on external contractors.

This lack of clarity at the leadership level can lead to AI investments falling short of expectations. A senior consultant from a large consulting firm, who wished to remain anonymous, stated that organizations are not achieving the anticipated return on investment and are struggling with employee engagement. This consultant's own firm provides employees with access to multiple AI tools, including specialized ones for tasks like coding, with some staff having access to four or five tools if their roles require it.

The human element of AI integration is crucial. "Organisations needed to consider the people side of the equation," the consultant emphasized, pointing to potential generational and gender differences in confidence levels with AI technologies. Before accessing any AI tool within his firm, employees must complete mandatory training covering AI ethics and risks such as bias. This training also educates users about the potential for AI tools to be sycophantic or to generate inaccurate information, often referred to as hallucination.