The success of artificial intelligence in the workplace will depend less on technical capability and more on how organisations support employees through change, according to employers, policymakers and wellbeing specialists meeting in Parliament this week.
At a roundtable convened by the Policy Liaison Group on Workplace Wellbeing, participants warned that without strong foundations of trust, culture and psychological safety, AI is more likely to generate anxiety and disengagement than sustainable productivity gains.
The discussion, held on Monday, brought together business leaders and experts alongside MPs including Sonia Kumar. Attendees agreed that wellbeing should be treated as core infrastructure for AI adoption rather than a secondary concern.
Evidence shared at the session suggested that organisations with strong, people-centred cultures are far more likely to implement AI successfully. Where trust is weak, training limited and employee support overlooked, new systems tend to provoke resistance and stress rather than improved performance.
Participants also cautioned against regulatory approaches that risk slowing innovation without addressing underlying organisational weaknesses. Instead, they argued for governance frameworks that help employers balance commercial objectives with employee needs, particularly in the context of the UK’s long-standing productivity challenge.
Several speakers highlighted the importance of creating environments in which staff can experiment safely with new tools, learn through practice and remove friction from routine work. AI is already being deployed in some organisations to support wellbeing directly, including through tools that help employees manage financial pressures or access support services discreetly.
Used responsibly, attendees said, such applications can enhance autonomy and reduce administrative burdens. However, these benefits depend on transparency, clear accountability and well-defined limits on how systems are used.
The roundtable concluded that treating wellbeing as an optional extra risks undermining the very gains AI is intended to deliver. Unlike earlier waves of workplace technology, AI systems evolve rapidly and place significant cognitive and emotional demands on users, making minimal training approaches ineffective.
While participants agreed that current applications are largely augmenting rather than replacing human work, they acknowledged that some roles will face significant disruption over time. This, they argued, makes early investment in reskilling and leadership capability essential.
Gethin Nadin, chair of the Policy Liaison Group on Workplace Wellbeing, said debates around AI had become polarised. “The conversation too often swings between hype and fear,” he said. “What’s missing is a serious focus on how technology interacts with human wellbeing. Wellbeing is not an add-on to AI adoption — it is a precondition for success.”
The group called on employers and policymakers to embed trust and psychological safety from the outset if AI is to improve job quality and performance rather than accelerate burnout.
