top of page

AI Reverse Delegation

The named move for routing a task upward to AI rather than downward to a team member, when handing it down would erode the team member’s self-determination.

​

Reverse Delegation routes a task upward to AI when the conventional downward route to a team member would tax the receiver while reclaiming little leverage for the leader. This was named and refined through ChangeSchool’s work with senior leaders across our executive education programmes, as part of the modular blocks of our The Art of Delegation curriculum.

​

The framework: the routing test and the Four Accountabilities

Two conditions, both of which must hold for the upward route.

​

reverse_delegation_routing

Reverse Delegation routing diagram: from a task at centre, three arrows show keep (leader does it), delegate down to a team member (the Charan reflex), and delegate up to AI (the reverse-delegation move). The routing test gates the choice; the Four Accountabilities wrap the upward arrow.

reverse delegation table

Both must hold. If only the first holds and the task is developmental, the right move is still downward, with the development made explicit. If only the second holds and the team member’s role is to do such tasks, the downward move is fine. Reverse Delegation applies only when both conditions are met.

 

When the routing is upward, the leader retains the Four Accountabilities.

​

Task definition. The leader specifies what the task is, what ‘good’ looks like, and what the output will be used for. AI does not define its own brief.

 

Process. The leader designs how the task will be done with AI: the interaction mode, the working relationship, the checks that run inside the work itself.

​

Outcomes. The leader is accountable for the output. If the AI-produced bibliography misattributes a quotation, that is on the leader. Routing transfers the doing, not the responsibility.

​

Quality improvement over time. The leader tracks how the task goes across instances and refines the brief, the process, and the checks. Each upward delegation gets better than the last because the leader is steering the iteration.

​

The Four Accountabilities are what distinguishes Reverse Delegation from abdication.

​

Genesis

William Oncken Jr. and Donald L. Wass (1974, Management Time: Who’s Got the Monkey?) named the original reverse-delegation problem: subordinates push tasks back up, and the manager’s calendar fills with other people’s monkeys. Peter Drucker (1967, The Effective Executive) sharpened the downward prescription, and Ram Charan, Stephen Drotter and James Noel’s The Leadership Pipeline (2001) operationalised it into the delegate relentlessly down training, with the rule of thumb of handing tasks to whichever team member can do them at 80% of the leader’s quality. The doctrine works because delegated work develops the receiver: 80% the first time, 90% by the third.

​

Edward Deci and Richard Ryan’s Self-Determination Theory (1985 onwards) identifies three core psychological needs that shape whether work feels meaningful: autonomy, competence, and relatedness. Phanish Puranam, in Re-Humanize (2024), reads SDT into the GenAI workplace. In practitioner shorthand, the three needs read as autonomy, mastery, and belonging.

​

Why it matters now

The Charan doctrine was built for a world where the rote tasks a leader handed down were also developmental. AI has unbundled those two things. The proposal-check is still rote; it is no longer developmental for the associate who has done three of them in six weeks. GenAI introduces a third routing option, and changes the trade-off underneath the downward one. The leader who applies the downward doctrine without checking what AI has unbundled erodes the team member’s mastery and belonging while reclaiming an hour of leader time.

​

The moves

Three moves install the routing test against the trained reflex.

​

Run the test on the next five delegations. Before handing any task downward, pause and ask the two conditions. If both yes, the task is a candidate for upward routing.

 

Pick one task that fails the test and route it up. The proposal-check, the glossary build, the bibliography compile. Write the brief, choose the interaction mode, own the output, plan how the brief will sharpen next time.

​

Tell the team member. Reverse Delegation is invisible by default. Name what AI is doing, what the leader is keeping accountable for, and how the output will be checked. Legibility converts a hidden routing decision into a development signal.

​

The discipline

Three habits embed Reverse Delegation as routine.

​

A weekly routing review of five minutes on a Friday: which downward delegations should have gone upward, which upward routings needed more rework than a downward route would have. The pattern recognition compounds.

​

A Four Accountabilities check before every upward delegation, written for that specific task. If they cannot be articulated, the routing is abdication.

​

A team-member development pass at the moment of routing: name what developmental work the team member will do with the reclaimed hour. Without this, freed time refills with the next rote task in the queue.

​

The discipline is to install a routing test that runs before the reflex, and to apply it where the underlying assumption of the downward doctrine has broken. Downward delegation remains the right move on most tasks; it is no longer the only move.

​

​

​

‘The leader’s reflex is to delegate downward; the discipline is to know when the right move is upward instead.’

 

 

Viren Lall, Managing Director,

ChangeSchool LDN (2026).

virenlall.com/ai-reverse-delegation

​

​

​

Ready for the Next Step.png

AI for Leaders.
Executive Education that changes practice.

Viren Lall is Managing Director of ChangeSchool LDN, a London-based executive education partner. ChangeSchool specialises in AI for senior-leader development, winning the EFMD Global Excellence in Practice Award in 2023 and 2025, with programmes in 39 countries.

​

Since April 2024, ChangeSchool LDN has been designing and delivering mindset shifts through Executive Education Programmes across sectors such as deep tech, manufacturing, and education, for business owners, governance professionals, and senior leaders. Leaders gain AI fluency, protect decision quality, spot value creation opportunities, and foster human-centric AI use. AI capability for senior leaders is also a core element and a constant spine of our Open Programmes for Chief Digital Officers, Chief Operating Officers, and Chief People Officers, delivered by our partner business schools.

 

Some of our clients include the Royal Academy of Engineering, Education and Training Foundation, and the UK Government's Meet Smart programme.

 

For speaking, programme, or partnership enquiries, get in touch with him through ChangeSchool LDN.

Get In Touch
bottom of page