Are You Developing Leaders for a Reality That No Longer Exists?

The real risk of agentic AI isn’t that machines will take over - it is quiet deskilling through convenience.

As agentic systems absorb the planning, synthesis, and administrative work that used to sit with leaders, the foundational human capabilities - judgement in ambiguity, navigating difficult conversations, and staying with complex problems - quietly atrophy.

The first documented case of AI-related deskilling emerged in 2025: clinicians working with AI-assisted detection saw their unassisted detection rates drop from 28.4% down to 22.4% once the tool was removed. (Budzyń et al., 2025). This demonstrates the dependence and vulnerability that can emerge when 'assistive' AI tools are introduced without deliberate safeguards.

This leaves Learning & Development (L&D) and executive teams facing two challenges operating in near-perfect tension:

1. Leadership development that ignores AI loses relevance.
2. Leadership development that defers to AI erodes the capabilities it exists to build.

When overreliance sets in, human leadership capability lapses. Yet, in an unproven and volatile AI economy where tool costs, availability, and performance remain uncertain, human judgement is not an overhead cost to be optimised away - it is your organisation’s ultimate economic insurance.

The LARM Diagnostic: Mapping Capabilities Across Realities

To ensure AI delegation is a deliberate design choice rather than an accidental drift, at LIW we’ve designed the Leadership-AI Relationship Model (LARM) which maps AI capability levels against human ownership.
‍

The gravitational pull of any organisation moves downward toward the Ceded row. Research on cognitive offloading shows that people naturally transfer cognitive work to external systems they find reliable, eroding the underlying skill over time

The Context Lenses

The same position on the matrix carries vastly different implications depending on your operational lens:

• Workflows: Shared + Agentic can be productive when co-producing work, provided both contributions stay visible.
• Leadership Decisions: Shared + Agentic introduces choice architecture risks, where AI subtly narrows options before human leadership takes place.
• Learning & Development: Shared + Agentic risks becoming "facilitation as theatre," where concentrating AI support on pivotal moments quietly strips away developmental friction everywhere else.
  ‍

Action: Map Your L&D Portfolio Against LARM

Most L&D teams discover a disconnect: they are designing for a fully human baseline while operating inside a shared human-AI reality.

To locate where capability currently sits across your organisation, a great place to start is to map each current L&D programme into the LARM matrix to pinpoint whether it reinforces Retained, Shared or Ceded ownership.
‍

Protect Your Leadership Pipeline

Don't let leadership capability risk fall to chance. Evaluating where your development strategy sits on the LARM matrix is the first step toward building an intentional integration model that protects human judgement.

👉 Download the full ‘The capability you should never hand to a machine’ Whitepaper to access the complete LARM diagnostic framework, detailed risk tripwires, and the 90-day action roadmap.