As we reach the end of 2025, it’s clear this year has reshaped the way organizations think about managing knowledge at scale. The pace of change has been relentless, driven largely by the explosive growth of AI and the shifting expectations that come with it.
What once felt experimental or “nice to have” has quickly become foundational, forcing teams across industries to rethink how they gather, verify, govern, and deliver information. In many ways, 2025 has served as a reality check - highlighting not just how powerful our tools have become, but how crucial it is to wield them responsibly, strategically, and in partnership with human expertise.
We knew going into 2025 that AI was going to be the word of the year, and wow, did it deliver. While we knew AI was trickling into people’s consciousness for the last couple of years, it has really mushroomed in the past 12-18 months, and the knowledge management industry has been no exception.
It may have seemed like jumping on the bandwagon in 2024, but this year, there was a change: AI stopped being a novelty and started becoming something organizations are considering on a strategic level.
Large language models now generate:
First-draft documentation
Contextualized summaries of meetings, projects, and customer interactions
Automated tagging and classification
The efficiency gains delivered to library teams and knowledge managers have become impossible to ignore. And we see AI as something that will soon become an essential part of every team’s plan to manage information at a large scale.
Here’s another trend that has been building for a while, but really took off in 2025. Information overload has been an increasing issue for years, with more data now posted online in a year than was generated in decades previously.
So what changed in 2025? Say it with me - AI! While people have been behind the boom in content in previous years, in 2025, we saw a huge uptick in the amount of content generated by Artificial Intelligence. This led to an increase in content with questionable trustworthiness - AI has known problems for hallucinating information, and people have had to become more likely to ask for sources or double-check the data. But now the problem is getting worse; AI has been found to hallucinate its source URLs as well.
And a report by OpenAI suggests the problem of hallucinations is getting worse with each update, not better.
While in previous years you might struggle to find the information you need, in 2025 you’ll definitely find it but you might not be able to trust that it’s true. Against this backdrop, finding sources that you know are high quality and reliable has never been more important.
Data governance is a growing challenge, and it’s even more important with libraries working at scale.
A key challenge of AI is that it pulls sources that it may not have legal access to, adding another layer of complexity for libraries trying to determine what they can and cannot share compliantly. Businesses need to know where the content came from, is it trustworthy, and do they have a right to use it. With generative AI producing so much content, the most mature organizations in 2025 track:
Source citations
Revision history and authorship
Model-generated vs human-authored content
Publisher reliability
Copyright
Without attribution, provenance, and quality markers, teams can worry whether or not they can trust the content sent out to them.
Library teams have historically been a little isolated from the rest of the company. Requests have been sent to them, and they have sent out the information required in alerts and newsletters.
However, with the influx of new legal technology, library and knowledge teams have the opportunity to step out of the shadows. As we have discussed, AI has had a transformative impact on information management, and library teams are well-positioned to take ownership of this new technology and align themselves with IT and technology teams within the business.
Integrating or merging with technology centres, and retaining control of key elements of new legal tech, put library teams in a much more integrated and dynamic position within the organization. And, crucially, one that often has bigger budgets and KPIs similarly linked to efficiency, automation, and reducing manual workflows.
A quick scan through this article and hundreds like it will show how much of a huge impact AI could have on the information management landscape. From summarization to categorization - even translation - AI can help with many of the manual, arduous tasks that currently make up a lot of the workload of library teams. Combined with automation technology, knowledge teams could see huge efficiencies by using their experience and skills and enhancing it with modern tech.
Efficiencies on scale work best when humans and modern tech work together - rather than one or the other taking the bulk of the workload.
2025 has made one thing clear: managing knowledge at scale now depends on pairing human expertise with AI-driven efficiency. This year pushed AI from “interesting experiment” to strategic necessity, while also exposing the risks of misinformation, content overload, and unclear provenance.
At the same time, knowledge and library teams stepped into a more integrated role - working closely with technology partners and taking ownership of how AI is adopted across the organization.
As we move forward, the message is simple: the future of knowledge management relies on balanced collaboration between people and technology, ensuring information is not only abundant but trustworthy and impactful.