EDRM 2.0 is here: What changed, why it changed, and what it means for your practice

EDRM 2.0 gives vendors new language for AI and disposition, but no new proof

EDRM has released EDRM 2.0, which it describes as the first substantive update to its reference model since it absorbed the Information Governance Reference Model. Roughly 150 practitioners worked on it for two years; the original dates from 2005. Nextpoint first published an interview on September 22, 2026, and EDRM republished it in October 2026. In it, EDRM's Mary Mack and Holley Robinson told Nextpoint's Amanda Fong about four changes. These are a unified data acquisition framework, governance as a foundation layer, analysis across every stage, and disposition as its own phase.

Is the acquisition change real, or just relabelling?

Mostly the model catching up. Mack says corporate legal teams increasingly work with data in place and do not collect immediately, and the interview says tooling and workflow had already moved.

Grouping identification, preservation, collection and processing therefore describes current practice rather than prescribing it. That helps when training junior staff, who meet the diagram first, but it won't change what your platform does. Governance gets similar treatment. It moves from off to one side to a base layer, on the argument that its principles should apply while discovery runs, not be checked once at the outset.

Does an AI layer in the model tell me anything about the tools I'm buying?

Little on its own. The model places analysis across every phase, but the published description is conceptual; nothing in the excerpt we reviewed says how an AI layer should be validated or audited.

Mack called analysis the "connective tissue" of the model. As a description of modern review that is fair. As a purchasing criterion it is empty. Expect vendors to claim alignment, and note that the excerpt mentions no certification or test. Ask each one which phases its AI touches, whether every answer traces to a source document, where the models run, and whether privileged material can ever train anything.

Why does the new disposition phase matter?

Disposition becomes an explicit step for handling data once a matter closes, and the interview singles it out as the addition with the most immediate practical consequences.

The excerpt we reviewed stops before the disposition discussion, so EDRM's specific advice isn't covered here. The tension is clear enough, though. The chain-of-custody records that protect you during a matter are the same ones you would need to justify a deletion afterwards. Ask whether your vendor can delete on a defined trigger, prove it, and say exactly what it keeps. Check, too, whether your hosting terms make retention the default.

How much weight does a reference model from the industry's own body deserve?

Treat it as shared vocabulary, not a standard. The excerpt cites no court or rule adopting it, and the roughly 150 practitioners behind it worked through a chartered consensus project.

Our source is Nextpoint, a vendor, interviewing EDRM's leaders. EDRM republished the piece and thanks Nextpoint as a Trusted Partner. The piece says the opinions are the author's, and it contains no independent critic. The 150-practitioner and two-year figures are EDRM's own. They measure effort, not outcomes.

Frequently asked questions

Is EDRM 2.0 mandatory?

The excerpt describes no obligation. It is a shared reference model, so adopt what helps your practice.

Does unified acquisition mean I can skip collection?

Not as a rule. The model groups the steps because teams increasingly work with data in place and do not need to collect right away.

Where does information governance sit now?

Underneath the whole model as a foundational layer, rather than upstream and off to one side.

Sources: EDRM, republishing Amanda Fong's interview first published by Nextpoint.