Skip to content
SHUMPEI KISHI
Go back

CMS Daishugo — Will Content Management Systems Survive the AI Era?

I recently attended an event in Fukuoka, Japan called “CMS Daishugo!” — which loosely translates to “CMS Grand Gathering.” It brought together communities from eight different content management systems: Drupal, WordPress, Plone, Craft CMS, Movable Type, microCMS, a-blog cms, and baserCMS — all under one roof to talk about the future.

As someone who’s spent most of their career in the Drupal ecosystem, hearing directly from developers and partners behind Python-based Plone, Craft CMS, and Japan-born platforms like microCMS and a-blog cms was genuinely eye-opening. I knew some of these names, but I’d never had a chance to hear their philosophies firsthand. It reminded me just how diverse the CMS landscape really is.

The main theme of the event was “AI and the Future of Content.” Two perspectives stood out to me in particular.

The first came from the keynote by Taku Fujita, CTO of Mitsue-Links. His talk was titled “Abandon CMS, or Evolve It?” — and he leaned heavily into the “abandon” side. He argued that AI’s evolution could make traditional CMS platforms obsolete altogether. Coming from someone who’s spent over 20 years working with every major CMS, those words hit hard.

The second was from Toru Kokubu, speaking about Craft CMS. His take was almost the opposite: CMS will continue to exist, but its role will shift. Rather than being a tool for creation, CMS becomes the “brake” to AI’s “accelerator” — a framework that defines what can’t be done, rather than what can. He described it as a “cage” for governance. That reframing stuck with me.

Across the sessions, one pattern kept emerging: MCP (Model Context Protocol) integration. The idea that AI tools like ChatGPT or Claude could operate a CMS directly through MCP was a recurring theme. If that trajectory continues, it’s not hard to imagine a future where people stop editing Word docs, spreadsheets, and CMS content themselves — instead, they just tell an AI what they want.

I’m already there, honestly. AI is an essential part of how I create slides and write documents today.

But then the question becomes: if AI does all the creating, do we still need the tools? Would Word and PowerPoint disappear?

I don’t think so — at least not yet. AI’s output still needs to be stored, shared, reviewed, and delivered somewhere. CMS, in that sense, still serves as the vessel — the container that holds and distributes what AI produces.

The real disappearing act would only happen when AI can handle the entire workflow autonomously — not just content creation, but delivery, review, permissions, and follow-up — without any human-facing interface at all.

That said, given how fast things are moving, I’d give it very low odds that I’ll hold the same opinion six months from now. For instance, “CMS is needed for content review and approval workflows” sounds reasonable today, but the moment AI can generate perfect previews and manage permissions on its own, that argument falls apart.

So what does a CMS need to survive? Honestly, I don’t think anyone knows. But if I had to pick one thing, it would be flexibility. The CMS that can adapt to changing demands, changing technology, and changing expectations is the one that will still matter.

Here’s where I’ll show my bias: I think Drupal is particularly well-positioned for this. Its modular architecture, robust APIs, granular permission system, deep customizability, and open-source nature make it a strong foundation for a world where AI and humans need to collaborate on the same system.

Imagine a CMS flexible enough to enforce strict human approval when needed, but also capable of letting AI agents handle everything from content creation to publishing to lead generation and follow-up. Is that even a CMS anymore? I’m not sure. But whatever you call it, I think that’s what the future demands.

CMS Daishugo!: https://cms.masizime.com


Share this post on:

Previous Post
The Efficiency Paradox
Next Post
Shakyo — The Day I Chose Not to Use AI