AI adoption targets need to reflect employee readiness

AI adoption depends on employee readiness

An AI usage target can make an implementation look successful before employees know how to produce useful work and take responsibility for it. When leaders set targets before understanding employees’ capabilities, support needs, and working conditions, recorded activity can reward the appearance of adoption. Employees may finish tasks faster while remaining unsure whether the results are accurate or appropriate. So managers responsible for an AI investment need to let readiness shape what they ask employees to do.

Employee reports reveal what usage counts leave hidden

The gap between activity and readiness appears in a recent survey by workforce analytics firm Visier. Visier has a commercial interest in employers investing in a better understanding of their workforce. Its findings describe employees’ reported experiences, while its recommendations address how employers should respond. Separating those perspectives helps leaders judge the findings and the advice that follows.

The findings on confidence show why recorded use can give managers an incomplete picture. In Visier’s survey, 48% of employees said they exaggerated their AI expertise at least some of the time, while 45% said they often felt pressure at work to use AI despite lacking confidence with the tools. Exaggerated expertise can lead managers to underestimate how much preparation employees need. Pressure can then push employees to participate before they are ready to benefit.

Visier describes participation without enough guidance or training as “AI compliance.” Employees may use newly introduced tools to meet an expectation while remaining unsure how to assess the results or choose suitable tasks. A rushed implementation can therefore produce visible activity alongside unresolved quality problems. Managers need to understand those conditions before treating a higher usage count as progress.

Uncertainty about the tools also sits alongside uncertainty about the organization’s plans. When Visier asked employees whether their organization had a clear AI implementation plan concerning its effects on employees and jobs, 59% said they did not believe it did or were unsure. Employees can therefore face an immediate expectation to use AI without a clear account of how their work will change. Explaining that change helps prepare them to participate.

Clear explanations also depend on leaders understanding employees’ current experience. Only 28% of Visier respondents felt their leaders fully understood how employees currently used AI, and 33% thought leaders understood their training and support needs. These responses point to a need for information to move both ways: leaders explain their plans, and employees describe their daily use and difficulties. An announcement can communicate a decision, but understanding its effects requires listening to the people expected to change their work.

Andrea Derler, principal researcher at Visier, connects this need for understanding to the choice of adoption measures. She warns that assessing individuals or teams mainly through usage percentages can give a misleading picture and focus attention on quantity. A count cannot explain whether employees chose suitable applications, protected information, or checked the resulting work. Those decisions require preparation that a usage target alone cannot provide.

Prepare employees to use AI safely and take responsibility

Preparation begins with an employee-adoption framework: an early plan for how people will learn to use AI, what rules apply, and how their roles may change. Derler recommends explaining how AI will be introduced, setting initial rules for its use, and describing the expected effects on employees. These steps give employees a basis for deciding how to work with the tools as implementation proceeds. They also clarify the responsibilities that come with each new expectation.

Those responsibilities start with handling information safely and understanding the tools available at work. “A few foundational skills and concepts every employee should understand include protecting sensitive data and uploading information with the greatest care, understanding the basic capabilities of approved tools available at work, and checking the work before sending AI-generated output to others,” Derler says. In practice, employees need guidance on general data security and recognizing false outputs. Before sharing work, they need to assess the information supplied to the tool, understand what it can do, and review its result.

Review needs a clear owner because the employee remains responsible for what is shared. “Most importantly, employees need to take full responsibility for the output. The AI isn’t responsible, the person using it is,” Derler says. An employer assigning that responsibility must also provide the knowledge and opportunity to exercise it. Otherwise, employees remain accountable for results they have been poorly prepared to assess.

Once employees can handle information and review results responsibly, they need to learn where AI is useful in their work. Preparation creates the conditions for experimentation: employees can try appropriate tasks, assess the output, and discuss what they find. Managers can use that experience to decide which uses to encourage. It also helps them distinguish useful applications from activity undertaken to satisfy a target.

Learn where AI fits before setting uniform targets

To build that experience, Derler recommends balancing required adoption with opportunities for employees to explore, discuss, and learn together. Employees can help identify activities AI can perform and activities it can support while a person continues to contribute. Their involvement gives leaders practical information about how work changes. Formal goals can then reflect what employees and managers have learned.

Derler explains why the sequence matters: “Most organizations are still learning which parts of a workflow AI can take over or augment, so creating opportunities for experimentation and learning is an important step before setting formal adoption goals or metrics,” she says. A workflow is the sequence of activities used to complete a task, and augmenting one of those activities means helping a person perform it. Experimentation lets employees and managers assess that division of work before making it routine.

A uniform percentage target can reverse this sequence by specifying how much AI to use before establishing where it belongs. Employees whose work has few appropriate AI applications face different conditions from colleagues with many suitable opportunities. Applying the same expectation to both groups can pressure people to find a use simply to meet the measure. Shared discussion helps managers account for those differences while leaving room for continued learning.

Setting those expectations also requires a sufficiently complete view of the workforce. “Many organizations, however, are still piecing together workforce data and don’t have a complete picture of their workforce, making it harder to understand where AI can create meaningful efficiencies,” Derler says. Employee feedback helps connect workforce information with the conditions under which tasks are performed. Mandates and metrics can then serve the organization’s needs while preserving opportunities to explore where AI is appropriate.

Connect training with changes in work and careers

Once leaders understand where AI could help, employees’ priorities can guide the support they provide. Asked what would improve their confidence in using AI at work, Visier respondents ranked training and upskilling opportunities first, greater leadership transparency second, career-growth and advancement opportunities third, and better communication about AI’s effects on employees fourth. Upskilling means developing additional skills for changing work. The ranking gives leaders a starting point for discussing practical preparation and employees’ longer-term concerns.

The career priorities extend the discussion beyond learning how to operate a tool. Employees want to understand how AI will change workflows and processes, particularly where existing skills and expertise may become irrelevant. A concrete picture of future work helps them adjust their expectations and career goals. Developing AI proficiency becomes a more meaningful commitment when employees understand what it could mean for their future at the company.

That connection to future opportunities also matters for employee engagement. When the organization’s direction is unclear, employees have a harder time deciding what to learn and how to prepare. Leaders need to connect changes in work with possible paths for growth and advancement. Employees can then make development decisions with a clearer understanding of why those capabilities matter.

Employers face financial choices alongside those development decisions. “The risk of underinvesting in training and upskilling is that the alternative can be much more expensive,” Derler says. She identifies the alternatives directly: “Layoffs, rehiring, and competing for top AI talent all come with significant direct and indirect costs. Investing in the workforce organizations already have can be an important part of making AI adoption sustainable.” Her recommendation puts training budgets alongside the costs of other workforce changes an employer might pursue.

The appropriate spending depends on what the organization wants to achieve with AI. Derler says some companies need stronger training and upskilling, while others may need additional resources, new talent, or new tools to help employees adjust to a changed pace of work and gain access to appropriate AI skills. For companies investing “millions” in AI, she argues that employee participation is essential to making that spending useful. Participation helps avoid performative use: activity aimed at appearing to adopt AI while leaving the value of the work unchanged.

Key takeaways for decision-makers

• Match adoption goals to employee readiness: Managers setting AI targets need to assess employees’ skills, support needs, and working conditions. Usage counts alone can reward activity while leaving quality problems unresolved.
• Make AI plans clear and listen to employees: In Visier’s survey, 59% of employees were unsure whether their organization had a clear plan for AI’s effects on employees and jobs or believed it lacked one. Executives overseeing implementation need to explain expected changes and gather feedback on daily use and training needs.
• Prepare employees to own AI output: Training needs to cover sensitive data, approved tools, false outputs, and review practices. Employees remain responsible for shared results and need the knowledge and time to check them.
• Test workflows before setting formal targets: Managers and employees need opportunities to identify tasks AI can perform or support. Use those findings to set expectations that reflect the work each team does.
• Connect training with career opportunities: Employees ranked training and upskilling as their top confidence-building priority. Workforce investment decisions need to connect skill development with future roles and weigh training costs against layoffs, rehiring, and competition for AI talent.