“Our people just resist change.”

I’ve lost count of how many times I’ve heard this explanation for stalled AI adoption. It’s rarely accurate and never useful.

Resistance isn’t random. It has causes. Different causes require different responses. “More change management” isn’t a strategy—it’s a hope.

Diagnosing the actual barriers to AI adoption allows targeted intervention. Let me share a diagnostic framework that works.

The Diagnostic Mindset

Effective diagnosis requires specific mindsets:

Curiosity Over Judgment

Resistance is data, not a character flaw. Approach it with genuine curiosity: What’s causing this? What would make sense of this behaviour?

Judging people as resistant closes inquiry. Curiosity opens it.

Systems Thinking

Resistance rarely has single causes. It emerges from systems—incentives, structures, cultures, histories. Look for systemic patterns, not just individual attitudes.

Specificity

“Resistance” is too vague to address. Get specific: Who is resistant? To what exactly? Under what circumstances? That specificity enables action.

Validation

Some resistance is reasonable. Concerns about job security aren’t irrational. Scepticism about overhyped technology isn’t unreasonable. Resistance might signal legitimate problems worth addressing.

The Six Barrier Categories

AI adoption barriers typically fall into six categories:

1. Capability Barriers

People lack the skills to use AI effectively:

Symptoms:

• People try AI tools but get poor results
• Frustration followed by abandonment
• Low confidence in ability to learn
• Avoidance of training opportunities

Diagnostic questions:

• Can people complete basic AI tasks successfully?
• What support have they received?
• What’s the learning curve been like?
• Where specifically do they struggle?

Root causes:

• Insufficient training
• Training that doesn’t match needs
• Inadequate practice time
• Lack of ongoing support

Interventions:

• Better designed training
• More practice opportunities
• Ongoing coaching and support
• Peer learning arrangements

2. Motivation Barriers

People don’t see reason to adopt AI:

Symptoms:

• Low urgency around AI adoption
• Completion of training without behaviour change
• “We don’t need that here” attitudes
• Preference for existing approaches

Diagnostic questions:

• Do people understand why AI matters for their work?
• What benefits would adoption provide them personally?
• What’s the cost of not adopting?
• What competing priorities exist?

Root causes:

• Unclear value proposition
• Benefits accrue to organisation, not individual
• No consequences for non-adoption
• Higher-priority demands on attention

Interventions:

• Clearer communication of personal benefits
• Demonstration of value through use cases
• Integration of AI adoption into expectations
• Removal or reduction of competing priorities

3. Fear Barriers

People are afraid of AI adoption:

Symptoms:

• Anxiety visible in discussions about AI
• Avoidance of AI topics
• Defensive reactions to AI suggestions
• Rumours and worst-case speculation

Diagnostic questions:

• What are people afraid of specifically?
• Is fear about job loss, skill inadequacy, or something else?
• How prevalent is fear across the population?
• What information would address concerns?

Root causes:

• Uncertainty about job impacts
• Worry about appearing incompetent
• Past negative experiences with technology
• Lack of psychological safety

Interventions:

• Direct, honest communication about workforce implications
• Creating safe learning environments
• Visible success stories from relatable peers
• Addressing specific fears with specific information

4. Environmental Barriers

The context doesn’t support adoption:

Symptoms:

• People express interest but don’t adopt
• Adoption happens then fades
• “I don’t have time” as common explanation
• Manager behaviours that contradict AI adoption

Diagnostic questions:

• Do people have time to learn and experiment?
• Do managers actually support AI adoption?
• Are the right tools accessible?
• Do processes accommodate AI use?

Root causes:

• Competing demands on time
• Manager signals against adoption
• Tool access problems
• Process barriers

Interventions:

• Protected time for learning
• Manager accountability for team adoption
• Simplified tool access
• Process modification to accommodate AI

5. Trust Barriers

People don’t trust AI outputs:

Symptoms:

• Heavy verification of AI outputs (or complete distrust)
• Preference to do work manually “to be sure”
• Concerns about AI accuracy and reliability
• Stories of AI failures circulating

Diagnostic questions:

• What experiences have shaped trust levels?
• Are trust concerns justified by actual AI performance?
• What would build appropriate trust?
• Is the issue over-trust or under-trust?

Root causes:

• Negative experiences with AI errors
• Lack of understanding of AI limitations
• Insufficient guidance on verification
• Cultural emphasis on error avoidance

Interventions:

• Education on AI capabilities and limitations
• Clear guidance on verification approaches
• Sharing of successful use cases
• Building verification into workflows

6. Identity Barriers

AI conflicts with professional identity:

Symptoms:

• Philosophical objections to AI use
• Pride in traditional skills
• “Real professionals don’t use AI” attitudes
• Identity-based resistance

Diagnostic questions:

• How do people define professional excellence?
• Does AI threaten that definition?
• What would preserve identity while enabling AI?
• Are certain groups more identity-protective?

Root causes:

• Professional identity built on skills AI replicates
• View of AI as “cheating” or shortcut
• Status attached to traditional methods
• Generational or craft identity

Interventions:

• Reframing AI as enhancing rather than replacing expertise
• Emphasising human judgment in AI-augmented work
• Role models who combine professional pride with AI use
• Discussion of evolving professional excellence

Diagnostic Methods

How do you determine which barriers apply?

Direct Inquiry

Ask people directly:

• What makes AI adoption difficult for you?
• What would make it easier?
• What concerns do you have?
• What would address those concerns?

Direct questions often get direct answers—if asked without judgment.

Observation

Watch what happens:

• Where do people struggle with AI tools?
• What’s the body language in AI training?
• How do managers talk about AI?
• What happens when AI is discussed informally?

Observation reveals what people might not articulate.

Data Analysis

Look at adoption data:

• Who is and isn’t adopting?
• What patterns exist across groups?
• Where does adoption start and stop?
• What correlates with adoption success?

Data reveals patterns across populations.

Environmental Scan

Assess the context:

• What are actual time constraints?
• What do manager behaviours signal?
• What processes would need to change?
• What competing initiatives exist?

Environmental factors explain much resistance.

Cultural Assessment

Understand cultural factors:

• What are valued behaviours?
• How is failure treated?
• What’s the history with technology adoption?
• What stories circulate about AI?

Culture shapes what’s possible.

From Diagnosis to Intervention

Diagnosis should drive targeted intervention:

Match Intervention to Barrier

• Capability barriers need skill building
• Motivation barriers need value demonstration and accountability
• Fear barriers need psychological safety and honest communication
• Environmental barriers need structural change
• Trust barriers need education and appropriate verification guidance
• Identity barriers need reframing and role models

Generic “change management” addresses none of these specifically.

Address Multiple Barriers

Resistance often stems from multiple sources. A person might have capability gaps AND fear AND environmental constraints.

Comprehensive intervention addresses the full barrier profile.

Prioritise High-Impact Barriers

Some barriers affect more people or create larger obstacles. Prioritise based on:

• Prevalence across population
• Magnitude of impact
• Addressability through intervention
• Importance to strategic objectives

Sequence Thoughtfully

Some barriers must be addressed before others can be:

• Environmental barriers often must be addressed before capability barriers (people need time to learn)
• Fear barriers often must be addressed before motivation barriers (fear blocks engagement with value propositions)

Sequence interventions for effectiveness.

The Ongoing Work

Diagnosis isn’t one-time. Barriers evolve:

• New barriers emerge as adoption progresses
• Barriers shift as some are addressed
• Different populations face different barriers
• External changes create new challenges

Build ongoing diagnostic capability into your adoption approach.

The organisations that succeed at AI adoption aren’t the ones with no resistance. They’re the ones that understand their specific barriers and address them systematically.

Diagnose before you prescribe.