AI transformation is a human change programme enabled by technology
Most AI initiatives do not fail because the technology is inadequate. They fail because organisations underestimate the human response to AI, writes James Howarth, Senior Manager at change management consultancy Nine Feet Tall.
Despite significant investment, many AI programmes stall after pilots, struggle to scale, or fail to influence real decisions. The usual diagnosis points to data quality, model accuracy, or tooling.
In reality, the deeper issue is trust. Employees hesitate to use systems they do not understand; leaders are reluctant to rely on outputs they cannot explain; and organisations lack the governance to embed AI confidently into everyday operations.
Avoiding implementation pitfalls
Building trust in AI systems is critical. Transparent communication, ethical practices, and clear governance are not optional enablers; they are foundational requirements, trust is built when employees feel safe, informed and their views respected. At the same time, organisations must invest in reskilling and upskilling to ensure employees are prepared for AI-enabled workflows and remain central to transformation.
Recent trends show resistance to AI is increasing, not fading. Concerns around workers being replaced by AI, deskilling, algorithmic bias, and black box decision making are now widespread. The rapid rise of generative AI has intensified anxiety, particularly in professional and knowledge-based roles and in many cases, AI has been deployed faster than the organisational change needed to support it, creating uncertainty and mistrust.
These reactions closely follow the change curve widely adapted for organisational change frameworks. Applied to AI adoption, the curve provides a tool not just for explaining resistance, but for preparing transformation leaders and organisations in advance.
In the early shock or denial stage, employees may underestimate the impact of AI or assume it will not affect their roles. Organisations that prepare well start communication early with a clear narrative, explaining why AI is being introduced, where it will be used, and what will not change. Failing to address these topics early erodes trust in the initiative from the start.
As AI becomes more visible, denial often gives way to fear and frustration. Employees worry about job security, performance measurement, and loss of autonomy. At this stage, trust is built through transparency and psychological safety.
Leaders must acknowledge concerns openly, explain system limitations, and make ethical commitments visible. Clear governance around data use, bias, and accountability reassures employees that AI is controlled and aligned with organisational values.
Getting the most out of AI
The next phase, exploration, is where adoption either accelerates or stalls. Employees begin testing AI tools and assessing their usefulness. Organisations that succeed here learn to create the safety to fail fast and invest in practical, role specific upskilling. Training should focus on how to interpret AI outputs, when to challenge them, and how to combine AI insights with human judgement.
AI will fundamentally change the way they work and the tasks they perform and involving employees in pilots and feedback loops reinforces ownership and confidence in the tooling and how they will perform their role in the future.
Acceptance and commitment occur when AI is embedded into the operating model. At this stage, trust is sustained through consistency. Governance defines where AI is advisory versus mandatory, who remains accountable for decisions, and how systems are monitored over time. AI becomes a normal, trusted part of how work gets done and employees are continuously supported and encouraged to develop their AI literacy.
Ultimately, AI transformation is a human change programme enabled by technology. By building a programme around a proven AI change management framework, organisations can anticipate resistance, design for trust, and ensure that AI enhances human capability rather than undermining it.
