Much of the debate about AI and work has traditionally focused on capability: what the technology can do, which tasks it can automate and how quickly organisations can adopt it. But as organisations move from adoption to embedding, and from pilots to integration, the most important questions are shifting from what AI can do to what organisations choose to do with it.
For senior leaders, those choices increasingly reach into the design of the organisation itself. What happens to the time saved by AI? How will people acquire expertise if technology takes over much of the work traditionally performed at the beginning of a career? How much authority over workers should be delegated to algorithms? Who will gain most from access to increasingly powerful tools? And how will the value created by AI ultimately be distributed?
We are already seeing evidence of this shift: the World Economic Forum’s 2026 Four Futures for Jobs report models four very different labour market scenarios for 2030, showing how the same technologies can produce very different outcomes depending on workforce readiness, investment choices and governance. Five emerging fault lines will be particularly important.
1. Augmentation vs work intensification
The promise of AI augmentation is well-known: routine tasks can be completed more quickly and employees freed to spend more time on complex, creative or higher-value activity.
The productivity opportunity is arriving, however, against a difficult baseline. Microsoft’s 2025 Work Trend Index found that 53% of leaders believed productivity needed to increase, while 80% of the global workforce reported lacking sufficient time or energy to complete their work.
That makes the question of what happens to the time AI saves particularly urgent: if a task that once took five hours can be completed in one, the remaining capacity could be used to improve quality, develop new services or invest in learning; however, it could equally translate into higher output expectations or lower headcount requirements.
Evidence suggests both dynamics may already be emerging. Deloitte’s Human Capital Trends report found that 77% of surveyed employees said AI had increased their workloads, while 61% expected it to contribute to burnout.
Augmentation and work intensification can therefore occur simultaneously. Today’s exceptional AI-assisted output can quickly become tomorrow’s expected output.
For senior leaders, the strategic question is simple: what are we actually seeking to achieve when we ask people to do more with AI? The answer will shape workload, job quality, engagement, retention and the sustainability of the operating model itself.
2. Automation vs apprenticeship
Perhaps the most important long-term workforce question is also one of the easiest to overlook. Some tasks do more than produce immediate economic value and teach people how to become experts.
Across law, accountancy, consulting, technology, financial services and many other knowledge-intensive sectors, early-career employees have traditionally learned through research, first drafts, checking, analysis, repetition and feedback. Much of this work is highly amenable to generative AI.
This creates a paradox: organisations may be able to automate junior work long before they can automate the senior expertise that junior work eventually produces.
The Open University’s 2026 Business Barometer found that 19% of employers had reduced recruitment into early-career roles. Among those hiring fewer early-career workers, 42% attributed the change to wider AI use or AI performing more entry-level tasks.
For organisations, the strategic risk goes beyond job displacement and extends to pipeline erosion: removing foundational tasks may generate efficiencies today while weakening the mechanisms through which future specialists acquire pattern recognition, contextual understanding, professional intuition and judgement.
That does not require preserving repetitive work indefinitely. It does, however require organisations to become more deliberate about apprenticeship: development that once happened almost incidentally through productive work may increasingly need to be consciously designed.
For employers, the challenge is to automate tasks without accidentally automating the pathways through which expertise is created.
3. Optimisation vs worker autonomy
AI is also changing its position within the employment relationship. Increasingly, AI systems are being used not only by employees, but to organise, assess and make decisions about employees themselves.
The use of algorithms to manage work is already widespread. ILO research published in 2026 found that 74% of surveyed mid-level managers across six high-income countries said their firms used at least one algorithmic management tool to instruct, monitor or evaluate employees.
These technologies can support recruitment, scheduling, task allocation, productivity measurement, performance evaluation and workforce planning. Their appeal is obvious: organisations can process information at scale, identify patterns quickly and allocate resources more efficiently.
Yet optimisation also raises questions about employee autonomy: the ILO found that 46% of workers whose productivity was monitored electronically ‘all the time’ said they worked too fast, compared with 15% among workers who were never electronically monitored.
For senior leaders, the question is where the boundary between algorithmic recommendation and human judgement should sit. The more consequential the decision, the more important transparency, privacy, fairness, accountability and contestability become.
AI is taking workplace technology into new territory by embedding AI systems within the processes through which employees are managed and work is organised. The debate may therefore increasingly move from how employees use algorithms to how algorithms shape the way work is carried out.
4. Skills democratisation vs a new AI divide
One of generative AI’s most exciting possibilities is the democratisation of expertise. Capabilities that once required specialist knowledge or support are becoming available to far larger numbers of people.
Yet access to those gains is highly uneven. PwC’s 2025 UK Workforce Hopes and Fears Survey found that only 15% of UK workers used generative AI daily. Usage was heavily concentrated among senior employees: almost 60% of senior executives used GenAI at least monthly, while between 68% and 79% of non-managers never used it at work.
The economic returns are uneven too. PwC’s Global AI Jobs Barometer found that workers with AI skills commanded an average wage premium of 56%, while the skills demanded by employers were changing 66% faster in occupations most exposed to AI.
This creates the risk of a two-speed workforce. Some employees may gain access to high-quality tools, strong data, autonomy and opportunities to experiment. Others may experience AI primarily through automation, monitoring or rising performance expectations.
For senior leaders, this makes AI capability a broader question than training people to use particular tools: the deeper issue is how widely the benefits of AI-enabled capability are distributed across the organisation.
The next competitive divide may be shaped less by which organisations have AI and more by which help employees use it effectively across the business.
5. Productivity vs participation
All four of these fault lines ultimately converge on a larger question: who benefits from the value AI creates?
If optimistic expectations around AI productivity prove correct, organisations may generate substantially more output with the same resources, or the same output with fewer resources. But there is no single destination for that productivity dividend: it could mean higher profits, increased investment, higher wages, better jobs, shorter working time or reduced headcount.
The early signals are revealing: World Economic Forum research found that around 45% of executives expected AI to increase profit margins, while only 12% expected it to lead to higher wages. Deloitte’s 2025 research similarly found that more than half of respondents regarded sharing the rewards created by AI with workers as very or critically important, yet 77% reported that their organisations were doing little about it.
How employees share in the benefits of AI may matter for performance as well as fairness.. If employees repeatedly hear that AI is transforming productivity while experiencing higher workloads, reduced headcount or greater insecurity, trust may erode. Other trajectories are possible: gains can be reinvested in skills, internal mobility, better jobs and more sustainable patterns of work.
The defining question of the next phase of workplace AI may therefore be less about whether organisations capture a productivity dividend, and more about what they choose to do with it.
Shaping the outcome
These five fault lines point towards a wider shift in the AI debate: AI can increase productivity while intensifying work; automate routine tasks while disrupting pathways into expertise; improve organisational decision-making while reducing worker autonomy; democratise capability while creating new inequalities; and generate value while leaving open the question of who participates in it.
The CIPD’s 2026 research captures the transition well: AI is currently reshaping work primarily through changes within jobs, including task composition, skills requirements and workload intensity. The significance of AI therefore extends far beyond a simple calculation of jobs created or jobs lost.
For senior leaders, there is a fundamental opportunity in that uncertainty. The direction of AI in the workplace, and of the future of work more broadly, is neither fixed nor linear. We are at a crossroads, and the choices organisations make now about how AI is deployed, how work is redesigned and how its benefits are shared will help determine which direction we take.
For boards and executive teams, the strategic conversation needs to move beyond where AI can be deployed to what kind of organisation is being built around it.
While many of those choices will look operational, the organisations best positioned for the future will understand that AI offers more than a new set of tools, as it presents a series of choices about what work should become, how opportunity should be created and how the value generated by technology should ultimately be shared.
