Treat “AI sounds robotic” as useful feedback—and prove that the writing assistant supports your team rather than replacing it

Rohan Hall’s approach to AI adoption starts with making complex AI ideas accessible and putting the technology in a clearly bounded role. When an employee says an AI writing assistant “sounds too robotic,” the literal criticism may be accurate, but it is rarely the whole concern. The employee may also be asking whether craft still matters, whether management plans to lower quality standards, or whether using the tool will help justify reducing the team later.

Do not answer those concerns with “AI is inevitable” or “everyone needs to adapt.” Those statements turn a writing-quality problem into a contest over authority. Instead, agree on the observable point: generic prompts and unedited drafts often produce generic prose. Then explain that the team’s expertise is what identifies weak reasoning, missing context, inappropriate tone and language that fails to sound like the organization.

Resistance takes different forms. An employee who feels diminished by an AI email tool needs a different conversation from someone rejecting every new system. If confidence is the issue, consider how to introduce an AI email assistant without making employees feel inadequate. If long-tenured employees associate new technology with replacement, use a slower approach to help experienced staff adopt AI without feeling displaced.

The fastest way to create fear is to introduce a writing assistant without boundaries. “Use AI more” leaves employees guessing whether it is optional, how their performance will be judged and which parts of their role management hopes to automate. Replace that ambiguity with a short written agreement covering the tool’s permitted tasks, human responsibilities and quality threshold.

The boundary should also identify sensitive work that requires extra care. A routine internal outline is not equivalent to a difficult employee message, a contractual statement or communication involving private information. Establish categories before the pilot so speed does not quietly become the only standard.

A controlled pilot is more persuasive than reassurance because it lets the team see where the assistant helps and where it creates extra work. Choose a recurring, low-risk task that employees already understand. Avoid beginning with the most creative, sensitive or identity-defining work; that makes the test feel like an attack on professional expertise rather than an attempt to reduce friction.

Design professionals may object that using generative tools is “cheating,” which is fundamentally a question about authorship and craft. That deserves a direct discussion about acceptable assistance, not a lecture about efficiency; see the approach to addressing the belief that AI tools undermine creative work. In operational teams, open hostility may already be damaging work. In that case, stabilize expectations and workload before following the framework for responding to AI fear when productivity is falling.

Participation must involve real decisions. Asking employees for feedback after leadership has already chosen the tool, workflow and success criteria feels performative. Invite them to identify unsuitable tasks, contribute examples of strong writing, test prompts and define the conditions under which a draft is unacceptable.

Create a small working agreement in the team’s own language. It should answer: What is the assistant for? When should it not be used? What information can be provided to it? Who verifies factual claims? Who owns the final voice? How should employees report a poor result or an unsafe workflow? Revise the agreement as the pilot reveals new issues.

Managers should also model the behavior they expect. Show an imperfect AI draft, identify what is wrong with it and revise it openly. That demonstration sends a stronger message than presenting polished output while concealing the human work required to make it useful.

Adoption rate alone is a poor measure. Employees can appear compliant while quietly duplicating the old process, correcting extensive errors or avoiding the tool whenever supervision is absent. Evaluate the complete workflow, including preparation, prompting, verification, rewriting and approval.

Do not reward employees for generating more text. Reward clearer thinking, better communication and responsible use of time. If the assistant creates bland drafts that require complete rewrites, narrow its role to outlining, summarizing or generating alternatives. If it repeatedly makes the same mistake, change the instructions or remove that use case from the workflow.

Some resistance reflects a broader lack of trust in leadership’s technology decisions. A facilitated conversation can help when managers and employees are talking past each other. You may want guidance on finding an AI speaker who can engage a skeptical team or, where job-security anxiety dominates, choosing a speaker who addresses AI replacement fears directly. A speaker can create shared vocabulary, but leadership still has to make credible commitments afterward.

Rohan Hall simplifies complex AI ideas to make them accessible. He is the founder and CEO of OceSha Ventures and its AI-first solutions, which builds and operates course creation, branded academies, AI assistants such as Lumi, and business intelligence for businesses and organizations. His work is relevant when a team needs both a clearer understanding of AI and practical systems with defined business purposes.

For organizations exploring AI-assisted education, OceSha AI’s course and academy creation platform represents one application of that work. Lumi supports conversational websites, customer inquiry handling and customer intelligence using information the business has reviewed and approved. If your next challenge is grounding an assistant in reliable organizational material, start with how to train an AI assistant on your business information.

Rohan also examines AI alongside other major technological shifts in The Convergence of AI and the Top 10 Emerging Technologies. That broader perspective is useful for leaders who want to understand why AI changes work without reducing the conversation to either panic or hype.

The essential leadership task remains the same regardless of the tool: make the purpose clear, specify where human judgment remains decisive and let evidence determine whether a workflow deserves to continue. To explore Rohan’s ventures, book, podcast and wider work, visit Rohan Hall’s home page.