The Real AI Challenge Is Not Creating More Content

Ask an AI assistant to draft a paragraph and it will usually produce something usable in seconds. Ask where the approved record of a product-related decision is stored, whether it is still valid, or who is responsible for updating it, and getting the correct answer is much more difficult.

Generating text is increasingly easy, but making fragmented, unevenly documented, and constantly changing information usable is still complicated. As Artificial Intelligence is rapidly transforming our work, new tools, processes, and ways of collaborating are emerging and raising important questions about how information work will evolve in the future.

Content generation matters – as do translation, localization, and productivity – but they are only part of the picture. A larger opportunity can be found in improving how information moves through work: how people find it, assess it, connect it to other information, and act on it. Technical communication has been dealing with information discoverability, structure, metadata, reuse, and governance for decades. AI does not currently replace these capabilities.
  

Where the Real Struggle Lies

AI can produce an answer. Organizations still need to know if it is the right answer, based on the right information, for the right situation.

The Meaningful Digital Evolution (MDE) project examines how AI is changing work practices, information flows, and collaboration. Working with partner companies from several sectors as well as tekom Europe and the Finnish Technical Communication Society STVY, the project combines surveys, interviews, workplace observations, and practical experiments to form a picture of today’s developments.

When MDE started working with companies in August 2025, the general public discussion around AI mainly focused on individual use, content creation, and productivity. However, when we conducted surveys and interviewed knowledge workers in the partner organizations about their concerns around AI, a different picture emerged. People described:

• Fragmented information scattered across systems and teams
• Gaps in documentation and knowledge
• Unclear ownership and responsibilities
• Difficulty seeing what is current, relevant, or missing
• Considerable effort spent checking whether information can be trusted

This reflects a broader shift in the business use of AI, suggesting that instead of selecting AI tools to support individual tasks, content creation, and enhancing productivity, the most valuable AI opportunities can be identified by considering workflow problems. A missing decision, an outdated file, or an undocumented responsibility will slow down a project regardless of how much content the organization has produced.
 

Change vs. Efficiency

The tekom survey from 2024 showed AI use across technical communication, especially in translation and localization, information research, text optimization, content generation, and quality assurance. More than 63% of respondents expected their work to change within the following two years. However, at the time around 77% saw little or no time savings from AI. Tasks, responsibilities, and quality checks may change. But MDE findings also suggest that the immediate effect of AI is not that the same work gets done faster, but that new work is created around review, verification, and integration. Even in areas where AI adoption is already mature – such as in localization – organizations are increasingly confronting broader questions around information quality, ownership, traceability, and trust. Many of these challenges will sound familiar to technical communicators.

In general, instead of faster production, tekom survey respondents wished for a better division of work. Almost 80% viewed the automation of routine tasks positively, because it could create more room for complex, creative, and strategic endeavors. Unfortunately, just giving individuals access to an AI assistant does not make that wish come true.
 

From Isolated Tools to Better Workflows

Organizations often begin their AI journey by comparing tools. Which assistant should we license? Which model is the best fit for our use case? Which feature can summarize a meeting? These are understandable questions, but not at this stage of the process.

A more useful starting point is to analyze the work itself. What is the purpose of the activity? Where does information come from? Who needs it? Who owns it? Which rules shape its use? Where do delays, uncertainty, and repeated checks occur? Only after answering these questions does it make sense to ask whether AI could help.

A process map can reveal steps and handovers, but it may not show why or where work repeatedly breaks down. Looking at the wider activity brings responsibilities, rules, tools, and persistent tensions into view. It also helps to prevent an AI system from automating a poorly understood process or producing more content for an environment that already suffers from information overload.
  

Three Information Problems, Three Prototype Solutions

The MDE project did not stop at the research. We also created three practical AI tool prototypes that address recurring information-related problems reported by organizations. While these tools serve different purposes, they share a common goal: helping people find, understand, and make better use of organizational knowledge.

1. Structurizer: Recovering Value From Legacy Content

Much organizational knowledge remains locked in PDF files that are difficult to update, reuse, or connect to modern content systems. Structurizer explores how AI can support the conversion of PDF-based material into structured formats. The aim is not only to generate new text, but to make existing information easier to maintain and reuse.

2. TellMeMore: Making Workplace Experience Visible

Organizations cannot improve work they do not understand. TellMeMore explores AI-supported interviews as a way to gather employee perspectives and identify recurring workplace issues. By helping to capture, organize, and analyze qualitative insights, the tool supports a deeper understanding of everyday work experiences and organizational challenges.

3. Knowledge Identifier: Showing Gaps, Risks, and Ownership

Projects often contain warning signs that remain hidden in everyday communication: missing information, unclear responsibilities, or issues that receive too little attention. Knowledge Identifier explores how AI can help make such gaps and risks visible, supporting oversight and decision-making.

Overall, organizations need better ways to work with the information they already have. They need to find it, understand it, verify it, connect it, and use it.
  

Why Technical Communicators Belong in This Conversation

Technical communicators have long worked on problems that now appear central to organizational AI. We structure information, clarify relationships, consider user needs, manage terminology and metadata, and pay attention to quality, traceability, and reuse. We understand that information is useful only when people can access it in the right form and context.

This expertise is relevant far beyond the production of manuals or help content. When AI is introduced into organizational workflows, someone still needs to ask: What sources should the system use? How is information kept current? How can a user judge the answer? Who is accountable when sources conflict? What happens when crucial knowledge isn’t documented?

These are communication, governance, and work-design questions. Technical communicators can help connect them. The opportunity for technical communication is not to produce more with AI but to help organizations build information environments in which AI can be useful.
  

Start With the Work, Not the Technology

The emerging lesson from the MDE project is straightforward: Begin by identifying what prevents people from doing their work well. First, examine the information flows, responsibilities, and recurring tensions around that work, and only then decide if and how AI is an appropriate form of support.

Sometimes the answer may be an AI application. Sometimes it may be clearer ownership, better metadata, a shared vocabulary, improved documentation practices, or a redesigned workflow. Often, it will be a combination of these.

The future of technical communication will certainly include AI-assisted content. However, it looks like our more consequential contribution remains elsewhere: helping organizations make information trustworthy, discoverable, connected, and usable. There is no shortage of content – the challenge is turning information into something people can confidently work with.
  

Practical Questions for Organizations

• What information do people repeatedly search for or verify?
• Where do knowledge gaps, conflicting versions, or unclear ownership slow work down?
• Which parts of the workflow need better visibility rather than more content?
• What must remain traceable and accountable when AI is involved?
• Could the problem be improved without AI, or does AI add a clear form of support?

Find further information in the tekom Europe Whitepaper "AI in Technical Communication".