Productivity Without Progress: The Organisational Reality Behind AI

PARC’s 2026 conference centred on a simple but often overlooked point: technology is a productivity tool, not a productivity strategy. While much of the external narrative assumes that AI will unlock efficiency gains almost by default, the discussion in the room was more grounded in organisational reality. The starting point was not how to scale AI, but to understand whether it is solving the right problem, for the right people in a way that improves outcomes. In several cases, organisations are still at an early stage, mapping where work breaks down and where friction emerges before deciding where technology might add value. 

This emphasis on friction rather than adoption exposed a familiar gap between theory and practice. At a macro level, the case for AI-enabled productivity is well established. Inside organisations, however, progress is uneven and often localised. Teams optimise individual metrics or automate discrete processes, but without addressing wider bottlenecks, overall productivity does not improve. As one panellist noted, it is entirely possible for parts of an organisation to become more productive without the organisation itself becoming more productive at all. In that sense, many of the challenges associated with AI feel less like something new and more like a continuation of long-standing issues around data quality, system design and organisational silos. 

For reward leaders, this raises a more fundamental question about where to place their attention. Much of the operational burden in reward remains process-heavy, creating a clear opportunity for automation and augmentation. Yet the discussion suggested that the greater value lies elsewhere. As technology takes on more of the transactional work, the role of reward shifts towards decision support, design and judgement. This includes helping managers navigate increasingly complex decisions on pay, performance and skills, often in real time. It also means engaging more directly with how work is structured and how value is defined, rather than relying on static benchmarks that quickly become outdated. 

The conversation on culture and capability was equally nuanced. There was broad agreement that upskilling matters, but also a recognition that focusing only on individual capability risks missing the point. Culture is not simply about whether people can use AI tools, but whether organisations create the conditions for them to be used well. This includes clarity on where AI should and should not be used, trust in how data is handled and a willingness to move beyond rigid processes. Notably, there was caution around incentivising AI adoption directly. The issue is not whether employees use these tools, but whether their use improves outcomes. 

Perhaps the most unresolved question relates to talent and long-term capability. While many were optimistic that AI will augment rather than replace roles, there was concern about how work is being reshaped at the entry level. If routine tasks disappear, so too may some of the learning opportunities that underpin early career development. At the same time, there were suggestions that AI may raise average performance while narrowing the space for deeper expertise and critical thinking. This creates a tension for organisations: how to capture short-term efficiency gains without undermining the development of future capability. 

Taken together, the day pointed to a more sober interpretation of AI’s role in productivity. The technology is advancing quickly and its potential is real, but the barriers to impact remain largely organisational. Productivity gains will depend less on the tools themselves and more on how organisations choose to redesign work, develop capability and align incentives. For reward leaders, that places them at the centre of the challenge, not at its periphery.