To win younger staff’s buy-in for your chosen AI tool, replace forced adoption with a fair, workflow-based pilot

This guidance from Rohan Hall addresses the leadership and workflow problem first: an AI product cannot create commitment simply because an agency owner selected it. Younger, digitally savvy employees may understand technology well enough to notice weaknesses quickly. Their resistance could reflect poor output quality, extra work, unclear data rules, an awkward fit with existing processes, fear of being judged by unfamiliar software or frustration that the decision was made without the people expected to use it.

Hold individual conversations before a group debate. Ask each employee to describe one task where the tool is useful, one where it fails and one condition that would make it worth using. Request examples rather than general reactions. “It is bad” should become a specific account of an incorrect output, duplicated step, slow approval or client risk. “Management never listens” should become a concrete decision the employee wants the team to revisit.

This diagnostic step matters across generations. The approach for employees who fear displacement is different from the approach for people who consider AI professionally illegitimate. Compare the guidance for helping veteran employees adopt AI without feeling replaced with the response to creative professionals who say using AI tools is cheating. Both groups may refuse the same technology, but their reasons call for different leadership responses.

If you introduced the tool as a finished decision, reopen the implementation process—not necessarily the purchasing decision. State what is fixed, what is open to change and why the agency is considering AI. Employees can help determine suitable tasks, review requirements, workflow placement and pilot measures. Leadership remains responsible for deciding which systems the agency will support, how work is governed and whether the investment continues.

That message avoids two bad extremes. The first is command-and-control adoption: “I bought it, so use it.” The second is leaderless consensus: “We will proceed only if everyone likes it.” Neither produces a sound operational decision. A better compact gives employees meaningful influence over how the tool enters their work while making participation in a legitimate evaluation part of the job.

Be equally clear about the business problem. “We need to use AI” is not a useful objective. Name the work that needs improvement: drafting routine material, organizing information, handling recurring questions or reducing repetitive administrative effort. If you cannot connect the product to a real bottleneck, do not blame the staff for failing to see its value.

Broad mandates create broad resistance. Choose one recurring task that is important enough to test seriously but contained enough to supervise. Map how the task works today, identify where the AI tool enters the process and preserve human review wherever accuracy, judgment, confidentiality or client impact requires it. The pilot should test the complete workflow, not an isolated demonstration produced under ideal conditions.

Do not use adoption rates alone as proof of value. Employees can comply with a mandate while quietly performing the real work elsewhere. Examine whether the AI-assisted output is usable, how much correction it needs and whether the total process is better than the previous one. For administrative technology in particular, refusal often persists when complexity shifts work rather than removing it; the same principle applies when staff say an AI scheduling system is too complicated.

Resistance often intensifies when rollout language makes capable people feel diminished. Avoid telling digitally savvy staff that the tool is “easy,” that younger employees should “get it” or that reluctance proves they are behind. Those statements convert a process disagreement into a status conflict. Instead, recognize that competent employees are precisely the people needed to evaluate output, identify failure patterns and decide where professional judgment belongs.

Training should begin with the actual workflow and expected standard, not a tour of every feature. Show what enters the system, what comes out, what must be checked and who remains accountable. Let staff practice on representative work where mistakes can be discussed safely. If employees feel embarrassed by the tool, use the same principles described for a small team whose AI assistant is making employees feel inadequate: lower the social threat while keeping the performance expectation clear.

Address job concerns directly. Explain which responsibilities remain human, which repetitive activities may change and how good judgment becomes more—not less—important when machine-generated work requires review. Do not promise that roles will never evolve. Give staff credible information about the current decision and invite questions about workload, standards and accountability. When a sales team believes automation is a replacement strategy, the relevant next step is a direct conversation about AI prospecting and fear of job loss, not another feature demonstration.

End the pilot with a decision meeting built around evidence. Ask four questions: Did the tool improve the chosen workflow? What new risks or burdens appeared? What changes would make the workflow acceptable? Is this product still the best available fit for the problem? Record the decision and explain it. A tool that does not earn a useful role should not survive merely because the owner chose it; a tool that does earn that role should not remain optional indefinitely because some people dislike change.

Once you decide to adopt, write a short operating standard. Specify the approved use case, required review, prohibited information, escalation path and owner of the workflow. Then set a date to revisit it. This prevents the agency from drifting into two unofficial systems: management’s declared process and the process employees actually use.

If the conflict reflects a wider generational divide, solve it as an organizational design problem rather than assigning one age group the role of innovator and the other the role of obstacle. The broader challenge of modernizing technology without dividing the workforce requires shared standards, reciprocal learning and consistent accountability. Likewise, when employees mock an error-prone system, investigate the evidence while correcting disrespectful conduct; the framework for responding when staff ridicule AI software separates valid product criticism from behavior that damages the team.

Rohan Hall’s work sits at the intersection of AI strategy, systems and practical implementation. His professional technology career began in 1984, and he has built AI and blockchain systems, led technology strategy and architecture, worked across the United States, Europe and Asia, and advised on emerging technologies. Readers can explore his current ventures, writing and broader work on Rohan Hall’s home page.

Rohan founded OceSha Ventures and its AI-first solutions. The company builds and operates solutions for businesses and organizations, including course creation, branded academies, AI assistants such as Lumi and business intelligence. Lumi supports conversational websites, customer inquiry handling, customer intelligence and answers grounded in information the business has reviewed and approved. These capabilities can support an AI initiative, but they do not replace the leadership work of selecting a real problem, involving employees and setting responsible operating rules.

For the broader technological context, Rohan is the published author of The Convergence of AI and the Top 10 Emerging Technologies. The book explores AI alongside major emerging technologies. That wider perspective is useful because workplace adoption is not just a software-selection exercise: leaders must understand how new systems change workflows, decisions and the distribution of expertise before asking teams to rely on them.