KMS was yesterday’s news, but AI made it strategic again
Inside Legal Platforms, by Bart Bogaerts
For years, knowledge management systems (KMS) sat quietly in the background of legal technology. Important, yes. Strategic? Not really. Many firms treated KMS as a useful archive: a place to store templates, precedents, and a bit of institutional memory. Necessary, but hardly exciting. I can tell you: that view is now outdated.
At ILTA last month, one thing became unmistakably clear: the market is no longer moving in small steps. It is shifting in layers, and those layers are starting to collide. In 2025, the legal industry learned how to prompt. In 2026, teams started building agents. The next phase is already visible: orchestration across systems, with platforms becoming more important than standalone tools.
That matters, because once everyone has access to the same large language models, the same legal drafting tools, and the same public legal content, the real differentiator is no longer the model. It is your context. And that context lives in your KMS.
The legal AI market is splitting into four layers
To understand why KMS suddenly matters again, it helps to look at the legal AI market as four interacting layers.
- At the bottom are the general LLM players: OpenAI, Anthropic, Mistral, DeepSeek and others. These companies provide the core intelligence layer. They are powerful, fast-moving, and increasingly commoditized.
- On top of that sit the legal AI application players: the Harvey, Legora, Saga type of tools. These are the company’s attracting attention and investment because they package AI into something lawyers can use more directly. Right now, they look exciting and some of them are already valued at extraordinary levels.
- Then there is the legal content layer: providers such as Thomson Reuters and Wolters Kluwer, who bring statutes, case law, commentary, and jurisdiction-specific content. That layer remains critical, because legal AI without trusted legal content quickly becomes shallow.
- And finally, there is the platform and knowledge layer: the systems that store the firm’s own legacy, templates, matter history, know-how, and company-specific context. This is where DMS and KMS come together. This is where firms store the things that make their work theirs.
That fourth layer is where long-term value sits
Because if every lawyer uses the same LLM, the same legal drafting assistant, and the same external legal content, then everyone starts getting similar answers. The only way to make output more relevant, more nuanced, and more firm-specific is to bring in proprietary knowledge. Not generic knowledge, but your knowledge. That is why KMS is no longer yesterday’s news but it is why it is becoming strategic infrastructure.
The hidden value of
“boring” systems
There is another reason that this matters. The legal AI market is full of what looks like sexy tooling. But a lot of that tooling is not sticky. A lawyer can use one drafting tool today and another one tomorrow. That can change surprisingly fast, but a DMS or KMS is different. These platforms hold the long-term memory of the organization. They store:
- documents;
- templates;
- work product;
- historical matters;
- internal knowledge and
- filing structures that firms depend on every day.
You do not casually replace that. And that makes the platform layer far more strategic than many firms have realized.
Put simply: Harvey may be fashionable, but your knowledge platform is harder to replace.
That is exactly why legal leaders should stop viewing knowledge management as a side function. In an AI world, KMS becomes one of the few ways to create differentiated output at scale. SharePoint is not dead, but its role is changing. Another misconception that keeps surfacing is the idea that SharePoint is somehow becoming less relevant. In fact, the opposite is happening.
SharePoint may be less visible as a user destination than it once was, but it is becoming more important as a platform. At ILTA, the real takeaway was not that SharePoint is disappearing. It is that its role is evolving. It is moving from “place where documents live” to middleware and intelligence backbone for Microsoft’s broader AI environment. That matters because agents, Copilot, Teams, search, and Microsoft Graph increasingly depend on it.
If SharePoint is the platform on which agents run, permissions are respected, context is assembled, and workflows are connected, then storing legal content outside that environment starts to look increasingly inefficient. Why place documents in a third-party silo, only to pull them back out again when you need them for AI, search, or orchestration?
Open ecosystems
will beat closed ones
Many legal organizations are reaching a decision point: some are still on-premises, others are rethinking older platform choices made before AI changed the rules. In both law firms and corporate legal departments, there is growing recognition that the future is not just about document storage. It is about cloud readiness, AI readiness, and platform flexibility. The difference between open and closed ecosystems becomes more critical.
A closed platform may offer attractive built-in features, but it can limit choice. An open Microsoft-based environment offers something different: the ability to work with Copilot, build or connect agents, retain control over data, and benefit from broader innovation happening across the Microsoft ecosystem. As AI adoption accelerates, that flexibility becomes more valuable.
This is especially relevant as corporate legal teams begin building their own AI agents. That trend is already visible. Corporate legal teams are using SharePoint Agent Builder workflows to save significant time and create faster above-average results. Law firms should treat that as a wake-up call: your clients are moving too.
“Turns out boring was valuable all along.”
A new job for legal platforms
The biggest lesson is this: legal platforms are no longer just repositories, but they are becoming coordination layers. The platform will increasingly decide:
- how knowledge is surfaced,
- how permissions are respected,
- how agents interact with content,
- how workflows connect across systems,
- and how future-proof your legal environment really is.
That is a very different role from the old “upload and search” model. It also means firms should stop asking whether knowledge management is still relevant. The better question is whether their current platform is ready to turn knowledge into action.
Because in an AI market full of noise, novelty, and new vendors every month, the organizations that win will not be the ones with the flashiest tools. They will be the ones with the strongest foundation.
And that foundation increasingly starts with the systems many people once thought was boring.
Turns out boring was valuable all along…
