Re-Focusing Leadership on AI Readiness & Enablement
If Enterprise AI's prize is smarter organisations, then we need leadership functions to engage deeply with AI readiness and steering, rather than just buying more technology
Why Focus on AI-Enabled Organisational Change Rather Than Just Technology?
As we look ahead to another year of rapid technology-driven change in business and society, it is a good moment to separate the wood from the trees and focus on medium-term goals.
Generative AI has come a long way, and provides some wonderful capabilities, but it also carries risks. Agentic AI is full of promise, and could provide a safer, more manageable architecture for using AI in the real world; but it is still relatively unproven and depends on lots of other moving parts to operate successfully, which are mostly not yet in place.
Ultimately, we see AI as a social technology not a magic black box, and we are trying to discern the outlines of AI-enabled organisations amidst all the hype, and design the practical transformation actions and journeys that will help get us there.
We are interested in better ways for people to work together to build capabilities and solve problems, and we believe the future of work can be more elevating and rewarding than the old industrial era model.
This is why we spent so much time last year helping leaders transform their organisations using the superpowers that AI technology provides to connect, collaborate, and coordinate work more fluidly than before, with less bureaucratic management methods. It is also why we see the potential of AI not just in terms of automating existing process work, but as a way to create smart structures and platforms on which people can focus on value creation rather than ‘busy work’.
Leaders as Programmers of the Organisation
Great organisations are like alchemists - they create something from nothing, making customers lives better whilst creating value for owners, investors, workers and partners. In the past, the best firms relied on visionary leaders brave enough to do things differently. But instead of building and running them from scratch every time, using only job roles, functions and department structures as templates, what if we could develop them like software, using existing libraries and building blocks to create our own organisational operating systems that use smart automation and coordination to obviate the need for manual management?
We began 2025 fired up by the goal of AI-enabled organisational improvement and the long-term goal of smart, programmable organisations that can achieve both agility and scale without bloat and waste.
But throughout the year, we were reminded how existing management and incentive structures continue to drive short-term visible actions at the expense of longer-term readiness, architecture and planning. The urge to be seen to do something - anything - in enterprise AI has led to license purchases with no adoption planning, innovation theatre that generates press releases rather than meaningful capabilities, and KPIs that are about counting trees, rather than thinking about what can be built with new wood.
We hope 2026 will see more focus from leaders on AI readiness and all the technical and non-technical enablers for agentic AI to realise its potential.
AI Readiness Priorities by Function
We wrote a lot last year about AI readiness, and the use cases, leadership techniques and capabilities that leaders at all levels can utilise to make the most of AI tools and systems in their organisations today.
One common theme was the idea of legibility - making the implicit explicit, and surfacing the norms, rules and unspoken parts of the work system so that we can evaluate and improve them, and codify them into rulesets for both AI agents and people to work together better.
So, before we throw more coal into the furnace, here is a breakdown of the top 3-4 things we suggested executives can do as part of their existing work to improve AI readiness in their domains, and guide the conversation with technical teams to ensure AI projects meet real business needs now and in the future.
CEO & COO: From Oversight to “World-Building”
The primary shift for senior leadership is moving away from viewing AI as a “tool to be purchased” and towards seeing it as an “environment to be designed.”
World-Building as Strategy: Leaders should move beyond incremental efficiency and focus on defining the digital environment where humans and agents interact.
Context Engineering: Start defining the “organisational OS” to ensure the right information and data are available to people and agents to support reliable ways of working.
Cultivate AI Leadership Skills: Thinking and writing more like architects or developers than bureaucrats to create the clarity needed to guide AI development.
CIO & CTO: Infrastructure for the Agentic Era
In 2025, CIOs moved from LLM experimentation to planning and building robust scaling layers and collaborative architectures, but readiness challenges remain.
AgentOps & Scaling: Building the organisational infrastructure for agent development, deployment, and monitoring to achieve practical impact.
Small Models, Local Context: Making more use of safer Small Language Models (SLMs) to keep AI closer to local context, culture, and control.
Collaborative Architectures: Creating the “Context Plumbing” that allows people and machines to share a common information landscape.
HR: Redesigning Work and Human-AI Teaming
HR is starting to focus on the future of human-AI collaboration and “Centaur” capabilities, plus they will need to play a role in making the implicit rules of work explicit.
Designing Centaur Teams: Reducing organisational drift by clearly defining how humans and AI agents can work together most productively.
Codifying the Invisible: Taking unspoken rules and cultural norms and turning them into explicit rulesets and guardrails for AI collaboration.
Work Designers: Encouraging a mindset shift where every employee becomes a “Work Designer” of composable workflows.
Learning & Development: Shared Practice & Transformation
L&D is evolving from learning content production and simple training to facilitate “co-op mode” learning, practical experimentation and always-on in-the-flow AI learning systems.
Learning in “Co-Op Mode”: Accelerating adoption through shared practice rather than individual use.
Building the Future One Agent at a Time: L&D teams exploring agentic tools to address learner needs while staying grounded in practice.
Avoiding “Workslop”: Moving away from low-quality AI filler by focusing on the “work” of organisational transformation.
Bringing it all Together
To bring this all together - and also bring it to life for busy leaders and leadership teams - requires the knowledge and experience to address specific local conditions and challenges. But we are also developing repeatable learning interventions and leadership techniques that people can pick up and run with.
If you would like to learn more about this work, please get in touch.
For now, here is one simple method to bring together function leads to build a common map that can guide AI-enabled capability development, and which can achieve results in a single 30-day sprint:
Strong framing on context engineering as the real work here. Most orgs are still treating AI like they treated cloud adoption ten years ago, thinking its jsut infrastructure to buy instead of a fundamental redesign of how work flows. The centaur teams concept nails the real challenge because it requires codifying all those implicit norms and workflows that people just know but never document. I've watched companies skip this step and wonder why their AI agents keep making nonsensical decisions, its always because nobody bothered to make the invisible visible first.
"Leaders as Programmers of the Organisation"
I like this.