Dana Aubin, Comtech Services
July 15, 2025
Conversations about generative AI (GenAI) often veer into extremes—either full-scale automation or cautious skepticism. But for those actively working in technical communication, the reality is more nuanced. During a recent CIDM roundtable discussion, a more balanced, grounded picture emerged: GenAI is not a replacement for expertise, but a tool that can benefit writers when used strategically. This article distills that conversation into key takeaways for leaders who are shaping how GenAI fits into documentation workflows.
GenAI as an outlining partner
One of the clearest messages in the conversation was that GenAI is most useful early in the content development process, especially during planning and outlining.
“It starts as generating outlines based on targeted keywords and the audiences,” shared one member. Writers are leveraging AI to map out potential topic areas and identify knowledge gaps before drafting begins. That draft outline is then reviewed and refined with input from subject matter experts (SMEs), with the writer maintaining ownership of the final structure.
The role of GenAI here is not to dictate what to write, but to act as an external perspective that challenges assumptions and prompts. It helps writers determine, “What pieces am I missing? Does this make sense for our product?” Asking these questions during outlining and drafting can save rework later in the content development process.
Key insight: GenAI is most valuable as a first step in creating quality content, not a one-stop solution. Used this way, it enhances strategic thinking without replacing it.
Mitigating knowledge bias, especially for lone writers
A recurring theme was the challenge of working in isolation. Several members noted that when you’re the only technical communicator on a project, it’s easy to overlook content gaps.
“I go back and write the content, and then do a final validation with AI to see what I missed. But I still send it to engineering for confirmation.”
“If you’re a lone writer, there’s a lot you don’t know you’re missing,” one member acknowledged. “There’s a certain amount of knowledge bias.”
In these cases, GenAI becomes a sort of virtual peer reviewer—surfacing ambiguities or prompting questions a typical user might have. It’s particularly useful when internal feedback loops are limited or time constrained.
Still, this doesn’t eliminate the need for SME collaboration. As one member put it, “I go back and write the content, and then do a final validation with AI to see what I missed. But I still send it to engineering for confirmation.”
Key insight: GenAI can help identify blind spots, but human review is still essential, particularly when documentation supports complex technical systems.
Validation is mandatory for GenAI-generated content
A significant part of the conversation centered on validation. GenAI often produces well-structured, fluent output, but that polish can mask inaccuracies.
“If I just paste an AI output and say, ‘Ship it,’ then we’re going to get burned.”
“If I just paste an AI output and say, ‘Ship it,’ then we’re going to get burned,” one member warned. Several attendees described layered review cycles where content written with GenAI assistance is reviewed by SMEs for technical accuracy and then reviewed by an editor or peer reviewer for tone, scope, and appropriateness for the intended audience.
Key insight: A formal review workflow is non-negotiable when GenAI is involved. Speed must not come at the cost of accuracy, quality, or audience trust.
GenAI cannot replace audience analysis
Technical communicators often serve multiple audiences, and content must be customized accordingly. This level of contextual adaptation remains beyond what GenAI can reliably do on its own.
“You can’t use the same messaging for developers as you would for end users,” one member explained. “Even if the product is the same, their context isn’t.”
“You can’t use the same messaging for developers as you would for end users,” one member explained. “Even if the product is the same, their context isn’t.”
While GenAI can draft general-purpose content, tailoring by persona remains a distinctly human task. Several members noted that content created using GenAI often lacks nuance, particularly when it comes to instructional tone or troubleshooting language.
Key insight: Audience analysis and persona-driven writing remain essential. GenAI can support content development, but human authors must ensure relevance.
Prompting Is a skillset
Another insight was the increasing importance of prompt engineering. Writers are not just generating content—they’re building, refining, and reusing prompts to direct GenAI output toward specific content goals.
“We treat prompts like templates. And we revise them over time,” one member shared. Others described collaborative approaches where teams maintain shared prompt libraries and review prompt strategies together.
Members agreed, prompting skills are a necessary part of the authoring toolkit. It requires understanding how GenAI models work, crafting clear input, and evaluating output with critical judgment.
Key insight: Treat prompt development as a professional skill. It is something to be practiced, refined, and shared by teams.
Governance sets the parameters for use
As GenAI adoption increases, so does the need for clear policy. Members emphasized the importance of documenting how, when, and where GenAI should be used within content development workflows.
“If you don’t have a policy for how AI is used in your content workflows, you’re already behind.”
“If you don’t have a policy for how AI is used in your content workflows, you’re already behind,” a member shared. Teams are now defining rules around source attribution, review requirements, data handling, and acceptable use cases. The purpose of governance is not only to protect against risk, but also to ensure consistency, accountability, and strategic alignment.
Key insight: GenAI use should be governed by documented policies that reflect the organization’s values and priorities, not just its tools.
Human oversight is still the core of quality
Throughout the conversation, one principle remained clear: GenAI is a tool, not a replacement for technical communicators. It offers efficiency and insight, but it cannot replace experience.
“AI can write words,” noted a member. “But it can’t replace relationships, domain knowledge, or judgment. That’s still on us.”
“AI can write words,” noted a member. “But it can’t replace relationships, domain knowledge, or judgment. That’s still on us.”
This perspective reinforces the role of technical communicators not only as writers, but as stewards of accuracy, clarity, and relevance. This role remains essential even when GenAI is involved in the content development process.
Key insight: The strategic value of technical communicators is their ability to evaluate, contextualize, and communicate complex information, not just generate it.
Conclusion: Use AI with purpose
The key takeaway is this: GenAI can be a powerful tool for content development when it’s used with intention. The benefits are real—faster outlines, early identification of content gaps, increased quality—but so are the risks if implementation outpaces oversight.
The organizations seeing the most value from integrating GenAI in the content development process are the ones applying it within thoughtful frameworks. They understand that technical communicators are key to content development. They understand the limits of GenAI. They maintain rigorous validation of the content; they support their teams with clear governance and professional development.