Deloitte just dropped its latest AI report.
The headlines focus on model spend, agentic systems, and the race to production. But the real story sits buried in the data — and it lands squarely on us.
AI isn’t failing because the technology isn’t ready. AI is failing because organizations haven’t changed how people work.
That single insight reshapes where change management fits in the enterprise today. Here are the five signals from the report that every change manager should pay attention to this week.
1. The real problem isn’t AI. It’s adoption
Deloitte found that access to AI tools expanded rapidly. Yet fewer than 60% of employees actually use them in their daily work.
Read that again.
Companies pay for licenses. They deploy platforms. They announce rollouts. And nearly half the workforce still sits on the sidelines.
For decades, leaders treated technology adoption as a distribution problem — buy the tool, train the users, measure logins. AI breaks that model.
The gap between access and activation is where change management earns its keep. Stop measuring rollout completion. Start measuring activation — how often people use AI to complete real tasks, and what outcomes that use delivers. Change management software makes that visibility possible at scale.
2. Most organizations sit stuck in pilot mode
Only a small share of companies have scaled AI into production. The rest experiment.
Pilots multiply. Production shrinks.
You know this pattern. It’s the classic change failure curve: lots of activity, very little sustained adoption.
Three things drive it:
- Leaders confuse experimentation with transformation
- Pilots live outside the operating rhythm of the business, so nobody owns the adoption plan when it is time to scale
- Scaling AI triggers a cascade of downstream changes in roles, processes, and governance that the pilot team never prepared for
Change management doesn’t support AI. Change management determines whether AI succeeds or stalls.
3. Companies train people — but they don’t redesign work
Here’s the finding that should frustrate every change manager:
84% of organizations have not redesigned roles or workflows around AI.
They hand employees powerful new tools. Then they ask them to work the same way they always have.
Upskilling matters. But upskilling alone cannot close the adoption gap. The bottleneck isn’t skill. It’s the shape of the work itself.
Real AI adoption requires two changes moving in parallel:
- Behavior change in how individuals decide and collaborate with AI
- Operating model change in how teams structure roles and flow work between humans and agents
Change managers own both. But you cannot redesign work at enterprise scale from a spreadsheet. Change managers who still run programs out of Excel will drown in the complexity of AI transformation. Change managers who use purpose-built change management software will lead it.
Want to see ChangePlan in action?
Book a 15-minute walkthrough and see how ChangePlan turns AI ambition into adoption.
Book a demo
4. Governance is really a people problem
As AI agents start acting autonomously, governance becomes critical. Yet only about 1 in 5 organizations have mature governance in place.
Most leaders read that and think: we need a better AI policy.
Wrong.
Governance isn’t a document. Governance is a set of human behaviors built on four pillars:
- Trust
- Accountability
- Decision-making
- Human oversight
Each one lives squarely inside change management. Policy writers cannot produce trust. Legal teams cannot produce accountability culture. Change managers can — if they have the tools to coordinate that work at scale.
5. The winners aren’t optimizing — they’re reinventing
Most companies use AI for efficiency. A smaller, more ambitious group uses AI to rethink how work gets done entirely.
That distinction matters.
Efficiency gains live inside the current operating model. Reinvention breaks the model and builds a new one. Everything shifts — roles, processes, metrics, culture — and every shift triggers its own adoption curve.
This is the shift that defines the next decade of our profession. Change management moves from supporting change to leading transformation. Change managers leading this kind of work think like transformation architects: designing the target operating model with business leaders, sequencing interventions, and holding the program accountable to adoption outcomes rather than milestones.
Where ChangePlan fits in
AI forces change management to scale.
Programs that used to touch 200 people now touch 20,000. Programs that used to span six months now run for years. Spreadsheets, email updates, and PowerPoint readouts cannot carry that weight.
This is exactly the gap ChangePlan fills.
ChangePlan is purpose-built change management software designed for the scale and complexity of modern AI transformation. Change managers use it to map stakeholders, sequence interventions, track adoption in real time, and report outcomes to executive sponsors.
The equation for leaders scaling AI is simple:
- Invest in the AI platform → unlock the capability
- Invest in ChangePlan → unlock the adoption
Without the second investment, the first underdelivers. The Deloitte data proves it.
The bottom line
Deloitte’s AI report reads like a technology story. It lands as a change management story.
AI does not reduce the need for change management. It amplifies it.
AI is not failing because the technology is not ready. AI is failing because organizations have not changed how people work.
That is exactly where change management steps in — and exactly where change management software like ChangePlan turns ambition into adoption.
