Artificial intelligence has accelerated how we access and process information, but crucially it has also changed how we judge whether it can be trusted—or not.
While AI has existed for decades, the mainstream adoption of generative AI over the past few years has altered both its accessibility and its impact. Capabilities that once required specialist tools are now available through simple, conversational interfaces, reshaping how information is requested, delivered, and evaluated.
For knowledge teams, this shift is already visible. Users increasingly expect faster answers, concise summaries, and tools that reduce the need to search across multiple systems.
At the same time, the reality of AI in practice is more complex. This article outlines a practical starting point: what AI can realistically do today, where its limitations remain, and how knowledge teams can engage with it effectively and responsibly.
The shift in how research happens
The most significant change is behavioral.
Traditional research followed a structured path. The researcher began by identifying the legal or knowledge concepts involved, developing appropriate terminology, and only then moving into searching, refining, reviewing, and validating results.
This process required active control at each stage, with understanding built incrementally through direct engagement with, and inspiration from, source material.
AI changes that dynamic. It allows the end user to move directly and rapidly from question to synthesized answer, often without engaging with the underlying sources. Prompting replaces searching, and synthesis replaces manual review.
This creates risk. The process becomes faster and more intuitive, but also less transparent. The challenge is no longer how to find information, but whether it is trustworthy and truly answers your question.
As a result, the role of the knowledge professional shifts from locating information to validating what has already been generated.
What AI can do for knowledge managers
The rapid expansion of AI tools has made evaluation the new core role for knowledge teams. The challenge is no longer identifying what exists, but determining which tools are credible, appropriate, and capable of delivering real value.
In this environment, knowledge professionals play a critical role. Their expertise in understanding the business, assessing sources, validating quality, and applying context extends naturally to the evaluation of AI systems.
A small number of use cases are already proving both practical and widely applicable.
Intelligent search and information synthesis
AI enables a shift from document retrieval to answer delivery. Rather than returning lists of results, systems interpret user intent and generate consolidated responses from multiple sources.
In practice, this allows insights from reports, emails, and knowledge repositories to be brought together into a single output, reducing time spent searching and enabling greater focus on analysis
Automated summarization and data extraction
AI tools are effective at condensing complex material into structured, accessible formats. Lengthy documents, transcripts, and datasets can be transformed into concise summaries or key-point extracts.
This supports faster understanding and more efficient interrogation of sources, allowing knowledge teams to deliver clear insights without extensive manual review.
Scalable knowledge capture
AI also enables more systematic capture of expertise. Content such as interviews, presentations, and internal communications can be processed into reusable knowledge assets.
In more advanced applications, this extends to structured knowledge harvesting, where expertise is converted into searchable resources that can be accessed across the organization.
What AI cannot do for knowledge managers (yet)
Despite its capabilities, AI has clear limitations. Understanding these is essential for maintaining trust in knowledge services and ensuring responsible use.
It cannot guarantee accuracy
AI systems are designed to produce outputs that are fluent, well-structured, and highly convincing, but not necessarily accurate. AI outputs must therefore be treated as a starting point, with verification remaining essential.
It cannot replace professional judgment
AI lacks the contextual awareness and domain expertise that underpin high-quality knowledge work. It cannot fully interpret nuance, assess risk in complex situations, or apply organizational context in a meaningful way.
This reinforces the role of the knowledge professional as an interpreter and advisor. Judgment, experience, and critical evaluation remain central to delivering reliable insight.
It cannot compensate for poor-quality data
AI depends entirely on the quality of the information it can access. If underlying content is outdated, incomplete, or poorly structured, AI will surface those weaknesses rather than resolve them.
In AI-driven environments, this creates a form of digital fragility. Content that is not clearly structured or machine-readable becomes difficult to retrieve and may effectively disappear from view. Maintaining a high-quality knowledge foundation is therefore essential.
Where should knowledge managers start with AI?
A focused, practical approach is more effective than attempting large-scale transformation.
At the same time, effective adoption requires more than implementation. It depends on developing an informed and current understanding of what is available. For knowledge professionals, this means staying current with emerging tools, approaches, and risks, through professional networks, industry events, vendor engagement, and continuous review of new developments.
This combination of practical action and informed awareness allows knowledge teams to make more confident decisions, avoid unnecessary experimentation, and guide their organizations toward sustainable use of AI.
Stabilize the knowledge foundation
AI readiness begins with the quality of the knowledge base. Content should be structured, current, and supported by consistent taxonomy and metadata.
This is familiar work for knowledge teams, but its importance is now amplified. A well-managed knowledge environment ensures that AI systems operate on reliable and meaningful information.
Focus on high-friction workflows
Teams should identify where time is currently being lost. This often includes searching for information, recreating existing work, or responding to routine queries.
A simple audit of these workflows can highlight where AI is most likely to deliver immediate value. Targeting high-friction activities allows teams to demonstrate impact quickly without significant investment or disruption.
Build AI fluency through practical use
Effective adoption depends on developing confidence as well as capability. Knowledge professionals need to understand how AI behaves, where it adds value, and where it requires oversight.
Hands-on experimentation with familiar content provides a practical way to build this understanding. Over time, this supports a more informed and critical approach to AI use in day-to-day work.
AI in Knowledge Work Augments and Enables
AI is reshaping knowledge work by accelerating how information is found, synthesized, and reused, but it does not replace the core principles of knowledge management. Its value lies in augmenting existing capabilities and enabling more effective use of knowledge.
To move beyond experimentation, organizations need to shift their focus from activity to outcomes. Metrics such as searches performed or content generated are not enough. What matters is impact: faster decision-making, higher quality insight, and, where relevant, the ability to generate value from knowledge.
At the same time, these capabilities reinforce the importance of professional judgment, content quality, and governance. AI can accelerate execution, but it does not determine what is correct, credible, or worth acting on. That responsibility remains with the knowledge professional and becomes more critical in an AI-enabled environment.
Future posts in this series will explore these themes in greater depth, including governance, evaluation of AI tools, and the evolving role of knowledge professionals as stewards of trusted information.
0 Comments