Every enterprise I talk to right now is asking the same question: "Are we ready for AI?" And almost every one of them is answering it wrong. They're auditing their data infrastructure, evaluating cloud capacity, and reviewing their tech stack. Important. Not sufficient. Because the organisations that will struggle with AI aren't the ones with messy data. They're the ones with cultures that can't absorb change.
What You Need to Know
- AI readiness assessments that focus only on technology miss the primary risk factors
- The three biggest blockers to enterprise AI adoption are cultural, not technical: change resistance, unclear ownership, and capability gaps
- Organisations with strong change management practices adopt AI faster, regardless of their technical maturity
- The readiness question isn't "can we run AI?" but "can our people and processes absorb it?"
78%
of enterprises say they're 'exploring AI' but only 15% have deployed anything
Source: McKinsey Global Survey on AI, 2023
2.5x
faster AI adoption in organisations with mature change management practices
Source: Prosci, 2022
The Technical Readiness Trap
I understand why organisations start with technical readiness. It's concrete. You can assess data quality. You can evaluate infrastructure. You can score cloud maturity. These things produce a nice RAG-status dashboard that executives can review.
But I've spent twenty years watching enterprise technology programmes, and the pattern is consistent: technical readiness is necessary but not predictive. Organisations with excellent technical infrastructure still fail at adoption. Organisations with mediocre infrastructure but strong change capability find ways to make it work.
The reason: AI changes how people work. And changing how people work is an organisational problem, not a technical one.
The Three Real Blockers
1. Change Resistance at the Middle Layer
Executive leadership is usually enthusiastic about AI. They've read the reports, attended the conferences, seen the demos. The team on the ground is often curious, sometimes anxious, but generally open if they can see how AI helps them specifically.
The resistance sits in the middle. Middle management. Team leads. Department heads. The people who are responsible for operational performance and who see AI as a risk to their current way of working.
This isn't irrational. Middle managers have built their careers on understanding how current processes work. AI threatens to change those processes, which threatens their expertise. If you're the person everyone turns to because you understand the claims reconciliation workflow inside out, a system that automates parts of that workflow is an existential threat to your organisational value.
Addressing this means giving middle managers a role in the AI transition. Not just informing them. Involving them. Their process knowledge is the most valuable input for designing AI systems that actually work.
2. Unclear Ownership
"Who owns AI in our organisation?" If the answer takes more than five seconds, you have an ownership problem.
In most enterprises right now, AI sits in a no-man's land between IT, innovation, and the business. IT thinks it's a technology initiative. Innovation thinks it's their mandate. The business thinks someone else should figure it out and then hand them a tool.
This ambiguity means nobody is accountable for AI outcomes. Pilots launch without business ownership. POCs are built by innovation teams with no path to production. The CTO approves infrastructure spend but the COO hasn't signed off on the process changes AI requires.
When I ask "who owns AI?" and three people answer at the same time, that's not enthusiasm. That's a warning sign.
Tim Hatherley-Greene
Chief Operating Officer
3. Capability Gaps You Can't Hire Your Way Out Of
The instinct is to hire. Data scientists, ML engineers, AI product managers. The problem: there aren't enough of them, they're expensive, and even if you get them, they can't deliver value without organisational capability around them.
What most enterprises actually need isn't AI specialists. It's people who can translate between AI capability and business problems. People who understand the operations well enough to identify where AI adds value, and who understand AI well enough to scope realistic applications.
These people already exist in most organisations. They're your best analysts, your sharpest process thinkers, your most adaptable team leads. They need upskilling, not replacing.
The Readiness Assessment That Works
I've started using a readiness framework that splits assessment into two halves: technical and organisational. Same weight. Same rigour.
Technical (the part you're probably already doing):
- Data accessibility and quality
- Infrastructure and cloud readiness
- Security and compliance posture
- Integration landscape
Organisational (the part you're probably skipping):
- Leadership alignment on AI's role and expected outcomes
- Middle management engagement and involvement
- Workforce capability and upskilling plan
- Change management maturity
- Clear ownership and accountability for AI outcomes
- Process readiness: which workflows are documented and ready for redesign?
Score each dimension honestly. The organisational scores will almost certainly be lower. That's the point. Now you know where to invest.
What to Do First
If you're an enterprise leader reading this and thinking "our organisational readiness is probably amber at best," here's where to start:
Appoint an owner. One person, with budget authority and executive backing, who is accountable for AI outcomes. Not AI technology. AI outcomes. Business results delivered through AI capability.
Engage middle management early. Before you pick a use case, before you evaluate vendors, before you build anything. Bring your middle managers in. Ask them what processes cause the most pain. Ask them where they spend time on work a system could do. Their answers will shape better AI applications than any top-down strategy.
Assess your change management capability honestly. If your organisation has a poor track record with technology adoption, AI will be no different. Invest in change management capacity before, not after, you start building.
Start small and visible. The best AI readiness programme isn't an assessment. It's a small win. One use case, well chosen, well implemented, with visible results. That win does more for organisational readiness than any assessment document.
AI readiness is a whole-of-organisation capability, not a technology checklist. The enterprises that will adopt AI successfully aren't the ones with the best data lakes. They're the ones that can change how they work, bring their people along, and give someone real ownership of the outcome.