UK 
United States
The arrival of AI in the workplace signals a momentous shift in how we work, think, and create value.
Once viewed as experimental, AI tools have swiftly become essential, transforming everyday tasks, boosting efficiency, and helping companies respond faster to customer needs.
Yet as the pressure to do more with less intensifies, many businesses may be questioning whether it’s appropriate, or even effective, to require employees to use AI in their day-to-day work.
The enforcement debate
Some organisations are mandating the use of generative AI tools to hit efficiency targets, drive consistency, or demonstrate a return on investment. On the surface, this approach appears pragmatic. Compulsory use can ensure alignment with company processes and deliver measurable short-term gains.
Beneath the surface, however, enforcing AI use within teams can risk something far more valuable than productivity metrics: trust. When employees feel forced rather than encouraged to use AI, the technology can quickly become a source of anxiety instead of empowerment. Fear of getting it wrong or simple resistance to imposed change can weaken both morale and innovation.
When employees feel forced rather than encouraged to use AI, the technology can quickly become a source of anxiety instead of empowerment
Over time, this can lead to ethical complacency, where people prioritise compliance over critical thinking, resulting in a disengaged workforce. This raises a crucial question for every organisation: can you truly harness AI’s full potential if your people use it out of obligation rather than choice?
Why ‘AI compliance’ alone doesn’t work
The appeal of automation is hard to resist, especially for leaders eager to demonstrate quick wins and immediate ROI. But introducing AI without first building employee confidence and competence can undermine even the most advanced systems.
CPO, PwC | Why we hosted our first-ever 'Learning Week'
A 2025 McKinsey report found that over one in five employees had received little or no support in learning how to work effectively with AI. When people don’t understand how AI systems operate, they’re more likely to take an AI-generated recommendation at face value rather than treating it as one input among many in complex decision-making.
Underuse can be just as damaging as misuse. Organisations often invest heavily in sophisticated platforms, only to see them gather digital dust as staff revert to manual processes. This not only wastes resources but also fuels the misconception that AI “doesn’t work,” when the real issue lies in culture and capability.
In insurance, for example, an algorithm might flag a claim as potentially fraudulent based on data trends. But only human expertise can determine whether that flag makes sense given the individual circumstances, customer history, and overall context of the situation.
