Leading Augmented Teams Reference Card

A one-page reference for leading a human-AI team day to day. Use it alongside the Augmented Team Leadership Diagnostic to focus your development.

Value comes from redesigning work around human-AI collaboration, not from automating tasks in isolation (McKinsey and Company, 2026). The leader's job is to allocate work deliberately, brief and verify like an agent boss, hold decision rights explicitly, and protect the human experience of augmentation.

The Task Allocation Frame

For each task, decide the split deliberately, reasoning about nuance, stakes, empathy and accountability (Raisch and Krakowski, 2021).

Delegate to AI. High-volume, well-specified work where output is easy to verify and stakes are low.

Share. AI drafts or analyses, the human shapes, judges and decides. Most knowledge work sits here.

Keep human. High nuance, high stakes, empathy or accountability that must stay with a person.

Redesign the Workflow (not just the task)

Map the workflow end to end, before and after
Find where steps disappear, merge or invert once AI is in the flow
Redesign with the team as a co-design, not an imposition
Pressure-test against the busiest week before scaling

Orchestrate and Verify (the agent boss)

Brief AI with explicit intent and a clear quality bar
Decompose complex work and coordinate across tools or agents
Calibrate trust case by case, not wholesale

Verification stakes ladder. Match checking to consequence: a quick glance for low stakes; a structured check for moderate stakes; independent re-derivation for high stakes. Avoid both over-reliance on fluent but wrong output and gains-cancelling over-checking.

Decision Rights and Accountability

Agree, in writing, where AI may act, where it recommends, and where humans decide
Name a single accountable human for every consequential outcome
Define the override route and the escalation route when output is questioned
Keep decision rights visible and in use, not documented once

Protect the Human Experience

Quantify released capacity and reinvest it into higher-value work
Update roles, objectives and development plans to the new shape of jobs
Monitor workload, skill maintenance, meaning and fairness
Watch for skill atrophy, hidden overwork and erosion of judgement, and act