The Fall of the Monolithic EDRMS

For more than three decades, the centralised Electronic Document and Records Management System — designed like a file compactus — has been the standard for enterprise information management. OpenText, IBM FileNet, Objective: not just software platforms, but institutional infrastructure. Then Gartner published its 2026 Magic Quadrant for Document Management, and the old order has collapsed.

The legacy incumbents have exited the leader’s quadrant entirely. Microsoft, M-Files, Box, Doxis and Hyland now occupy the top positions — all cloud-native, some fully metadata-driven, AI-ready and AI-enabling. They think about records management in a fundamentally different way.

Read between the lines of the 2026 report and the message is stark: the traditional EDRMS model has reached structural end-of-life. Gartner explicitly notes that 70–90% of enterprise data is unstructured, and that document management platforms are now “critical to enterprise application strategies that require AI-ready, unstructured data.”

That framing — from records custody to information fuel for AI — is a profound reorientation. The vendors who couldn’t make that shift (OpenText retreated to Visionary; IBM FileNet to Niche Player) share a common design flaw: they were built as digital representations of a paper file compactus. Static. Centralised. User-dependent. Catastrophically ill-suited to the AI era.

The Gartner signal is the what. But there is a more urgent why — one that the Magic Quadrant cannot tell you.

Every new SharePoint site, every additional SaaS application, every AI assistant deployed on ungoverned content widens the gap between technology investment and actual productivity. Organisations that deploy AI on top of fragmented, ungoverned information infrastructure will not become more intelligent. They will become faster at being wrong — faster at losing provenance, faster at breaching obligations, faster at eroding trust.

The solution is not swapping one EDRMS for another. It is a design philosophy shift — toward what we call an IRKM Services Fabric: interoperable, loosely coupled services that govern, classify and activate information wherever it lives, without moving content and without burdening users with manual classification.

The leading indicator of this shift is Zero Touch Records Management (ZTRM) — records automatically captured, classified, secured and disposed of through embedded AI rules. Cadmus has already deployed ZTRM in production using M-Files integrated with Microsoft 365, auto-classifying SharePoint content as records and managing them in-place. The vision is not theoretical. It is deployable today.

1. Clarify Problem Statement — Reframe from “EDRMS replacement” to business outcomes: faster knowledge retrieval, fewer GRC incidents, AI readiness.
2. Select Experimental Use Case — Contract lifecycle, case management, regulatory compliance — high-value, low-regret starting points.
3. Map Your Information Landscape — A “lite scan” to identify high-risk record classes and knowledge-intensive functions. You cannot govern what you cannot see.
4. Establish an IRKM Innovation Sandbox — Test Zero Touch and in-place management patterns in real tools — M365, SharePoint, line-of-business apps.
5. Prove Value, Then Scale — Build a benefits case and define a roadmap through reusable service components — not new point solutions.

The competitive divide in the AI era will not be between organisations that adopt AI and those that do not. It will be between those that build coherent, governed information fabrics beneath their AI — and those that don’t.

Information, Records & Knowledge Management: Solutions for the AI Era — our new white paper — delivers the complete framework: paradigm analysis, vendor assessment methodology, use case patterns, and a practical transformation roadmap. Essential reading for CIOs, Enterprise Architects, Records Managers and GRC leaders.