The Prosci ADKAR model is a goal-oriented change management framework that describes five sequential outcomes an individual must achieve for change to be successful: Awareness, Desire, Knowledge, Ability, and Reinforcement. Originally developed for general organizational change, ADKAR is particularly relevant to AI adoption because AI transformation is fundamentally a people challenge. The technology works; the difficulty is getting people to trust it, use it, and change their workflows around it.

The Five Elements Applied to AI

Awareness

Individuals must understand why AI is being introduced and what problems it is intended to solve. In AI adoption, awareness gaps are common because AI capabilities are often poorly understood. Employees may not know what AI can and cannot do, may conflate consumer AI experiences with enterprise AI, or may have concerns shaped by media narratives about AI replacing jobs. Building awareness requires clear communication about the specific AI use cases being deployed, the business reasons behind them, and the expected impact on individual roles.

Effective awareness activities include town halls with live demonstrations, role-specific impact assessments shared transparently, and honest communication about what will and will not change.

Desire

Awareness alone does not create adoption. Individuals must want to participate in the change. Desire is influenced by personal motivation, organizational context, and the perceived risk of adopting or resisting the change. In AI adoption, desire is often undermined by fear of job displacement, skepticism about AI reliability, or frustration with past technology rollouts that created more work rather than less.

Building desire requires addressing concerns directly, involving employees in the design of AI-augmented workflows, and ensuring that early adopters experience genuine benefits. Peer advocacy from respected colleagues who have used the AI system successfully is more effective than executive mandates.

Knowledge

Once individuals want to adopt the change, they need to know how. For AI adoption, knowledge goes beyond technical training on a new tool. It includes understanding how to interpret AI outputs, when to trust and when to override AI recommendations, how to provide feedback that improves the system, and how their role changes in an AI-augmented workflow.

Knowledge-building activities include hands-on workshops with real data, role-specific playbooks, and mentoring programs that pair experienced users with new adopters. Generic AI literacy training is less effective than training tailored to specific job functions and use cases.

Ability

Knowledge and ability are distinct. An employee may understand how the AI system works but struggle to apply that knowledge in practice, especially under time pressure or in complex situations. Ability develops through practice, coaching, and the removal of barriers that prevent effective use.

Common ability barriers in AI adoption include poor system integration that forces users into manual workarounds, insufficient computing resources, lack of access to necessary data, and organizational processes that have not been updated to accommodate AI-augmented decision-making. Addressing ability gaps requires observation of actual usage, identification of friction points, and iterative improvements to both the technology and the surrounding processes.

Reinforcement

Sustainable adoption requires reinforcement mechanisms that prevent individuals from reverting to old ways of working. For AI systems, reinforcement is especially important because trust in AI builds slowly and can be destroyed quickly by a single high-visibility failure.

Reinforcement activities include celebrating adoption milestones, sharing success stories and measurable outcomes, incorporating AI usage into performance evaluations, maintaining feedback channels for ongoing concerns, and continuously improving the AI system based on user input.

Sequencing Matters

ADKAR’s value is in its sequential nature. Organizations that jump to Knowledge (training) without first building Awareness and Desire find that training is poorly attended or quickly forgotten. Organizations that deploy AI systems without ensuring Ability find that usage drops after the initial novelty wears off. Each element must be substantially achieved before the next can succeed.

Measuring Progress

Each ADKAR element can be assessed on a simple scale through surveys, interviews, or observation. Scores below a threshold at any stage indicate a barrier point that must be addressed before proceeding. This diagnostic capability makes ADKAR practical for AI program managers who need to identify where adoption is stalling and allocate resources accordingly.