Leading Human-AI Teams: Decision Rights, Oversight and Work Redesign Training Course

Programme Summary

Leading human-AI teams is a five-day training course for senior executives at strategic level, culminating in a human-AI work design and oversight plan for a unit they lead. Copilots and AI agents are entering daily work faster than roles, controls and accountabilities are being redesigned. The approach helps leaders allocate tasks between people and AI, set oversight, calibrate trust and measure hybrid team performance. Mars Training Centre UK delivers this programme in leading human-AI teams.

The Business Case

Organisations that simply add AI tools to existing jobs rarely see the productivity they expect. Staff over-rely on outputs they cannot check or ignore tools they distrust, and accountability blurs. Leaders who redesign work deliberately, decide where humans must stay in control and track how hybrid teams perform turn AI investment into dependable results.

• Productivity and Cycle Time: Improve as routine tasks are automated and skilled staff are redeployed to judgement-heavy work.
• Decision Quality and Error Rates: Strengthen as oversight points, escalation rules and verification steps are built into workflows.
• Workforce Confidence and Retention: Rise as roles, skills paths and accountability are made clear to people working alongside AI.

Principles and Evidence Base

• Parasuraman, Sheridan and Wickens Levels of Automation Model: Structures the decision on how much of each task to automate in Session 1.
• Shneiderman's Human-Centered AI Framework: Guides the design of systems that combine high automation with high human control in Session 2.
• ISO/IEC 42001 Human Oversight Requirements: Frame the oversight roles, controls and records set for AI-supported decisions in Session 2.
• OECD AI Principles: Provide the human-centred values and accountability tests applied to team and role redesign in Sessions 3 and 4.

Intended Participants

• Executives accountable for the performance of units where AI tools and agents are being introduced.
• Leaders who sponsor transformation programmes that change roles and workflows.
• Heads of operations responsible for service quality, productivity and risk in automated processes.

The programme serves government entities, financial services, healthcare providers and energy and industrial companies.

Learning Outcomes

1. Participants will be able to diagnose which tasks in a workflow should be automated, augmented or kept human.
2. Participants will be able to design human-in-the-loop oversight with clear decision rights and escalation rules.
3. Participants will be able to evaluate risks of automation bias and over-reliance, and set measures to calibrate trust.
4. Participants will be able to construct new role profiles and team structures around copilots and AI agents.
5. Participants will be able to justify a scorecard for measuring hybrid team performance.
6. Participants will be able to prioritise change actions for a unit moving to human-AI teaming.

Programme Structure

Day 1: Leading Human-AI Teams and Task Allocation

Which work should machines do, which should people keep, and where should they combine?

• Hybrid Human-AI Working Concepts and Terminology
• Levels of Automation Model for Task Decisions
• Workflow Decomposition and Task Inventory
• Automate, Augment or Keep Human Decision Matrix
• Capabilities and Limits of Copilots and AI Agents

Day 2: Oversight, Decision Rights and Accountability

Who decides, who checks and who answers when AI contributes to a decision?

• Human-in-the-Loop, On-the-Loop and Out-of-the-Loop Designs
• Decision Rights Matrix for AI-Supported Decisions
• Human-Centered AI Design Principles
• ISO/IEC 42001 Human Oversight Controls
• Escalation, Override and Incident Rules

Day 3: Trust Calibration and Human Judgement

How are people helped to rely on AI exactly as much as it deserves?

• Automation Bias and Algorithm Aversion
• Explainability and Confidence Signals for Users
• Verification Routines and Sampling Checks
• Cognitive Load and Skill Fade Risks
• Psychological Safety for Challenging AI Outputs

Day 4: Redesigning Roles, Teams and Performance

How are jobs, teams and measures rebuilt around AI agents and copilots?

• Role Redesign and New Hybrid Job Profiles
• Team Structures With AI Agents and Human Supervisors
• Reskilling Paths and Capability Building
• Hybrid Team Performance Scorecard
• OECD AI Principles in Workforce Decisions

Day 5: Applying Human-AI Team Leadership

How are allocation, oversight, trust and role decisions combined into one plan?

• Task Allocation Review of a Participant Unit
• Oversight and Decision Rights Design for the Unit
• Change and Communication Plan for Staff
• Peer Challenge of Work Design Plans
• Human-AI Work Design and Oversight Plan Presentation

Methodology

The programme is built around case analysis and structured exercises. Participants examine how banks, hospitals, contact centres and public agencies reorganised work around AI and test oversight and role choices on a workflow from their own unit.

• Suggested activity: Classifying the tasks of a claims-handling workflow as automate, augment or keep human.
• Suggested activity: Building a decision rights matrix for an AI-supported credit or triage decision.
• Suggested activity: Designing a scorecard for a team that works with an AI agent.

Capstone Output

Each participant completes a human-AI work design and oversight plan for a unit they lead. It contains a task allocation map, a decision rights and escalation matrix, trust and verification measures, revised role profiles, a hybrid team scorecard and a change plan, ready for executive approval.

Frequently Asked Questions

Is leading human-AI teams suitable for executives without a technical background?

Yes. It suits senior executives, heads of operations and transformation leaders. It assumes leadership experience but no programming or data science knowledge.

Which models underpin leading human-AI teams?

It draws on the Parasuraman, Sheridan and Wickens levels of automation model, Shneiderman's Human-Centered AI framework, ISO/IEC 42001 human oversight requirements and the OECD AI Principles.

How is the human-AI teams work design and oversight plan used at work?

You take back a plan your executive team can approve, setting which tasks change, who oversees AI outputs and how hybrid performance is measured.

How does leading human-AI teams differ from general AI strategy programmes?

It focuses on how people and AI share work inside teams, including oversight, trust and role design, rather than enterprise AI investment or technology choices.