AI Is Rewriting the Rules of Records Management — and This Article Knows It
Zasio's 2026 trends roundup gets more right than most: it's practitioner-led, specific, and honest about the uncomfortable work ahead for the RIM profession.
Every January, the information governance space produces its share of trend forecasts. Most are variations on a familiar script: AI is coming, privacy regulations are multiplying, and organizations need to "do better" at managing information. Zasio's 2026 edition, published at the year's open, is more than that — and worth your time for reasons both obvious and subtle.
What the article is
This is a multi-contributor expert roundup from Zasio Enterprises, a firm that operates across both records management software and consulting. Five contributors weigh in, including a general counsel, senior consultants, a VP of Product, and a VP of Technology, each offering a distinct lens on what 2026 will demand of the profession. It runs long enough to be substantive without turning into a white paper.
“Archivists will have to rethink the concept of provenance — the history of records and their movements — in light of AI deepfakes.”
That line, buried in a bullet point from VP of Technology Warren Bean, is one of the most quietly significant observations in the piece. It hints at something the article handles well: the recognition that AI doesn’t just create new compliance burdens — it destabilizes foundational concepts the profession has relied on for decades.
What it gets right
The strongest contribution is from General Counsel Will Fletcher, who introduces the “records of compliance” framework. This category now must encompass AI impact assessments, data protection impact assessments, deletion logs for AI inputs and outputs, and documentation of the governance policies themselves. This is a genuinely useful framework, not a buzzword repackage. It operationalizes AI governance in terms that records managers can act on.
Rick Surber's section on information security is bracingly practical. Rather than treating AI-powered phishing as an abstract future threat, he grounds it in the concrete argument that strong records management, including clear retention rules, well-defined response processes, and continuity planning, functions as organizational resilience infrastructure. This framing elevates records management from administrative housekeeping to risk management, which is exactly the repositioning the profession needs.
Where it could go further
The article’s weaknesses are largely sins of omission. The ISO standards landscape, particularly ISO 15489 and the emerging ISO/TR work around AI-generated records, goes unmentioned. For an audience of practitioners who will need to translate these trends into auditable standards-based programs, that is a gap. Similarly, the international dimension is missing. GDPR enforcement trends in Europe and their interaction with AI documentation obligations would have considerably strengthened Fletcher’s compliance framework.
Stephanie Martin’s contribution on AI adoption barriers, including cost, accuracy concerns, and the primacy of data hygiene, is the most practically grounded section. Still, it feels slightly rushed compared to the others. The observation that poor metadata and disorganized records will undermine AI ROI deserves a fuller treatment. It is arguably the most actionable point in the piece for organizations at the start of their AI journey.
The bigger picture
What this article captures, perhaps more successfully than it realizes, is the profession at an inflection point. Records management has long operated as a compliance function, important but reactive, often underfunded, and culturally peripheral. The AI moment offers a genuine opportunity to reframe the discipline as foundational to organizational intelligence. That shift requires exactly the kind of thinking on display here, treating records not as artifacts of past transactions but as infrastructure for future decisions.
For practitioners, this article is a useful orientation document at the start of 2026. It will not give you a standards checklist or a vendor comparison, but it will give you language and a community of peers thinking rigorously about the same pressures you are facing.
I think if all the files for the various agencies and departments in our state archive and how useful they would be to reconstruct someone’s life as a genealogist…and I would love everything to be digitized so AI could access it.