AI is reshaping roles. Are leaders shaping the direction?
People are already reshaping their work through AI, often before roles and responsibilities have been formally reviewed. Recent research raises a timely leadership question: how do we connect that individual initiative with clear direction, thoughtful use of released capacity and shared accountability?
People are reshaping their roles. Where is leadership?
When we talk about AI adoption, the conversation often centres on tools, training and time saved. Yet something more fundamental is happening inside organisations: people are changing how they work, which tasks they take on and where they believe they can add value. The role described in a job description may already be some distance from the role being performed.
Three recent pieces of research have brought this into sharper focus. They raise questions about how leaders understand the changes happening around them, and how individual initiative connects to organisational purpose.
Work is changing from within
SAP’s latest analysis describes employees using AI to offload tasks while expanding the scope of their work, often without meaningful guidance from their managers or organisation. Its broader research programme includes 2,718 employees, 500 C-suite leaders and 60 HR executives, alongside other research components. The findings about employees changing their jobs draw on related work-redesign research, so those headline numbers should not be treated as the sample for every claim.
If people are already making decisions about how their work changes, what connects those decisions across the business?
Someone close to the work may spot an opportunity long before a leadership team does. That initiative deserves attention. It also creates a need for shared understanding. A change that helps one person can affect another person’s workload, the quality of a handover or who is accountable for a decision. Several useful individual changes may still leave gaps between teams.
For me, this makes listening to what people are actually doing an essential starting point for role review. Leaders need to understand which tasks have changed, what has become easier, what now requires more checking and where responsibilities have begun to shift.
What fills the space AI creates?
Researchers from LSE and Protiviti, writing in California Management Review, describe a follow-up survey of 859 workers. Training was more clearly associated with regular AI use than with reported time savings. Regular users were also more likely to report reinvesting released capacity in higher-value work.
That moves the discussion beyond whether someone has learned to use a tool. What do they now have the capacity to do, and has anyone agreed where that capacity would be most valuable?
AI can create space for deeper thinking, better questions and more considered judgement when it is used thoughtfully. These studies do not demonstrate that AI itself improves critical thinking. That remains a possibility to design for and evaluate. Faster completion can also lead to more tasks being added, more output to review or greater pressure to keep pace.
Leaders have a part to play in making room for judgement. If thoughtful work matters, it needs to be visible in priorities, expectations and the way contribution is recognised. Otherwise, released capacity may simply disappear into an already busy day.
Who has the opportunity to reshape their work?
Gallup’s American Job Quality Study adds an important perspective. Its 2026 survey included 15,482 US employees, with findings about AI’s reported benefits drawn from 7,845 who had used it at work. Regular use was more common among managers, graduates and people already holding better-quality jobs. Across the full employee sample, 52% wanted more influence over the adoption of new technology.
It would be easy to describe confident adopters as the organisation’s high flyers. But these findings do not establish that they are stronger performers. Access, autonomy, job type and opportunity may all help explain who can use AI and benefit from it.
That matters when leaders draw conclusions about readiness for change. People doing essential operational work may have fewer opportunities to experiment, less access to relevant tools or less permission to alter established processes. Their knowledge of the work is still valuable. An organisation reviewing its roles needs to hear from those people too.
Questions for Reflection
“Where has AI already changed what people do, beyond what their job descriptions capture?”
“What should released capacity make possible, and how will we know it has added value?
“Are responsibilities and decision boundaries clear when people change their workflows?”
"Whose experience is missing from our conversations about AI adoption?"
"How are leaders changing their own work, and what example does that set?"
Do your leaders know how people’s roles are already changing and have you agreed what those changes should achieve?
Closing Thought
Connecting initiative with direction
These sources use different methods and populations, and much of the evidence is self-reported. They offer useful signals rather than a universal prescription. My interpretation is that leadership needs to bring individual learning and organisational direction into a more deliberate conversation.
That means reviewing work from the ground up, while leaders also examine their own roles and decisions. Where can teams act independently? Where is a shared standard needed? Which decisions require human judgement, and who remains accountable when AI contributes to the work?
Clear boundaries can give people confidence to experiment. A shared purpose can help them judge which changes are worthwhile. Together, they create the conditions for initiative to strengthen the whole organisation.
Research and Sources
Sources checked on 9 October 2026. Publication dates and research fieldwork dates are distinct; check for updates before reusing findings.
SAP, 7 October 2026: Spoiler Alert: The Future of Work Is Already Here. Corporate research and analysis drawing on several components; the full underlying methods have not been assessed for this article.
California Management Review, 2 October 2026: Why AI Training Is Driving Adoption but Not Sustainable Productivity, Hitoshi Nishimura, Daniel Jolles, Grace Lordan and Fintan Canavan. Insight article reporting survey associations; full follow-up sampling details are not established here.
Gallup, 6 October 2026: AI Benefits at Work Unevenly Distributed, Jeffrey M. Jones. Employee survey conducted 26 January–24 March 2026; reported benefits are employees’ assessments, rather than independently measured improvements.
Tracy Filler is a leadership and organisational adviser who works with owners, boards and senior leaders, particularly in founder-led and generational businesses, when growth, succession or change demands a different way of leading. Through organisational discovery, strategic advisory and executive coaching, she helps leaders see the dependencies, behaviours and patterns that can be difficult to recognise from inside the business. Her approach creates space for honest thinking, challenge without judgement and a practical route from insight to action. She holds up the mirror and offers the map.
